DoctorPopi
  • Joined on 2023-06-06
DoctorPopi commented on issue mrq/ai-voice-cloning#435 2023-11-02 16:58:26 +00:00
[Discussion] What is a large dataset?

`That is just really small, the ljspeech will have probably 5-10k clips (for cca 20hours of audio). Tortoise model is quite big so it would be quite prone to overfitting with small number of…

DoctorPopi commented on issue mrq/ai-voice-cloning#435 2023-11-02 13:20:18 +00:00
[Discussion] What is a large dataset?

Yes, exactly. But it's not my tool. It's made by @mrq

I'm so sorry, I read the message too quick and I assumed it was mrq who had answered ^^'. My apologies

`Try to use really low lr rate…

DoctorPopi commented on issue mrq/ai-voice-cloning#435 2023-11-01 22:43:25 +00:00
[Discussion] What is a large dataset?

Hey! Thank you, yes indeed I'm just trying to refine an english speaking model for a game character. Thank you for the finetuned model, if I understand correctly, James Becker took the feminine…

DoctorPopi commented on issue mrq/ai-voice-cloning#400 2023-10-31 17:59:06 +00:00
Upscaled output creates bad quality?

Hey, I have another question concerning this matter : can you point in which script the conversion to 22050 Hz happens? Which method is used to do so, and finally, does the conversion happen if…

DoctorPopi opened issue mrq/ai-voice-cloning#435 2023-10-31 11:16:10 +00:00
[Discussion] What is a large dataset?
DoctorPopi commented on issue mrq/ai-voice-cloning#248 2023-10-31 11:04:24 +00:00
Overfitting with large datasets

Hello, I'm following up on this issue, because I'm still struggling with overfitting problems. Only, I've realized, they are not only occurring with large datasets, but also with smaller ones (200…

DoctorPopi commented on issue mrq/ai-voice-cloning#432 2023-10-30 12:15:21 +00:00
XTTS trainer has been released

Same here, Coqui TTS documentation is so damn hard to understand...

DoctorPopi commented on issue mrq/ai-voice-cloning#418 2023-10-20 18:19:53 +00:00
Two trainings with exact same parameters result in different curves

Hey MRQ, thank you for pinpointing this in the DLAS output file, I found it in the log indeed! I'm going to try that.

I'm not sure I understand why you don't think it's worth it though?…

DoctorPopi commented on issue mrq/ai-voice-cloning#418 2023-10-20 08:22:45 +00:00
Two trainings with exact same parameters result in different curves

Hey @epp thank you for your answer!

I don't see the seed parameter in the previous documents, can you help me pinpoint it?

DoctorPopi opened issue mrq/ai-voice-cloning#418 2023-10-19 15:23:54 +00:00
Two trainings with exact same parameters result in different curves
DoctorPopi commented on issue mrq/ai-voice-cloning#417 2023-10-18 15:58:48 +00:00
Line delimiter is heard in the output

Ah! I see! I'll try and dig into the code to find where it is used, see if there are some parameters or threshold I can tamper with... Thank you!

DoctorPopi opened issue mrq/ai-voice-cloning#417 2023-10-17 10:13:40 +00:00
Line delimiter is heard in the output
DoctorPopi commented on issue mrq/ai-voice-cloning#410 2023-10-12 13:22:14 +00:00
Learning state restart question

Noted, thank you :) I'm going to keep this thread open for now, if that's no trouble, in case I run into more questions. Have a great day!

DoctorPopi opened issue mrq/ai-voice-cloning#410 2023-10-12 12:57:11 +00:00
Learning state restart question
DoctorPopi closed issue mrq/ai-voice-cloning#409 2023-10-12 09:25:30 +00:00
Fresh install, no module named dlas
DoctorPopi commented on issue mrq/ai-voice-cloning#409 2023-10-12 09:25:26 +00:00
Fresh install, no module named dlas

EDIT : Solved

=> update conda and pip => install Ubuntu Build essentials (don't know if that really played a part in it but well...) => Re run setup_cuda.sh AND pip install -r requirements.txt…

DoctorPopi opened issue mrq/ai-voice-cloning#409 2023-10-12 06:27:48 +00:00
Fresh install, no module named dlas
DoctorPopi commented on issue mrq/ai-voice-cloning#325 2023-10-11 15:26:13 +00:00
Weird sound after most lines

What seemed to work for me was reset all parameters of generation to default, and progressively retweak them to reach my previous configuration. We’ll see whether it’s a long term solution!

DoctorPopi commented on issue mrq/ai-voice-cloning#407 2023-10-11 12:15:39 +00:00
OSError when running start.sh after fresh install

Worked like a charm for me, thank you! (I'm on Python 3.10.13 though)

DoctorPopi commented on issue mrq/ai-voice-cloning#248 2023-10-01 19:00:49 +00:00
Overfitting with large datasets

Hey!

Mmmmh indeed, that's very weird that you have a different pitch but otherwise good results. From my experience, it happened that I had to use another batch of the same voice (but not…