DoctorPopi
  • Joined on 2023-06-06
DoctorPopi commented on issue mrq/ai-voice-cloning#484 2024-04-28 15:03:40 +00:00
So is this abandoned?

This repo is indeed the best voice cloning tools I've tested for nearly a year, nothing except Eleven Labs gets better results than this...

Thank you @mrq for all the hard work and for pointing…

DoctorPopi commented on issue mrq/ai-voice-cloning#482 2024-03-27 11:26:53 +00:00
ModuleNotFoundError: No module named 'tortoise.utils.device'

Mmh...Strange! Have you ran pip install -r requirements.txt after the setup cuda? Maybe check that the installed version of Tortoise on your environment matches the one in requirements.txt?

DoctorPopi commented on issue mrq/ai-voice-cloning#480 2024-03-27 11:20:10 +00:00
Missing whisper and tensor error

About the reshape tensor error, do you have this during the generation? I found that it might be a sample problem (you must have at least as much samples as the number of generations you make),…

DoctorPopi commented on issue mrq/ai-voice-cloning#478 2024-03-27 11:18:04 +00:00
Noise at the end of generated voice

Hey! Have you checked whether you have a blank line in your prompt? I mean something like this: "The first line of your prompt [an empty space here]"

I've noted that you need to make sure…

DoctorPopi opened issue mrq/ai-voice-cloning#479 2024-03-17 16:36:22 +00:00
Sharing my audio dataset manager
DoctorPopi commented on issue mrq/ai-voice-cloning#473 2024-03-01 15:05:02 +00:00
metavoicio / metavoice-1B-v0.1

Interesting! I still have way better results with MRQ and my trained models (I have a challenging speaker though, with lots of nuances in his voice), but it's still better than all the open source…

DoctorPopi commented on issue mrq/ai-voice-cloning#458 2024-01-26 10:11:46 +00:00
ModuleNotFoundError: No module named 'torch'

Hey! Are you sure it's not a conflict of environment? Like you're on the right environment when launching your script in VScode, but when you run the start.bat you run it from the base environment…

DoctorPopi commented on issue mrq/ai-voice-cloning#464 2024-01-26 10:10:16 +00:00
torch.cuda.OutOfMemoryError: CUDA out of memory. OOM

Hey! From what I see in your logs there might be a problem of environment variable.

`Training] [2023-12-17T22:33:24.078703] CUDA_SETUP: WARNING! libcudart.so not found in any environmental…

DoctorPopi opened issue mrq/ai-voice-cloning#468 2024-01-25 20:00:42 +00:00
[Question] Segments : what are their role in the training?
DoctorPopi commented on issue mrq/ai-voice-cloning#459 2023-12-07 07:26:51 +00:00
help installing tortoisee

Hey,

There might be some answers here or here.

On which OS are you? Windows?

DoctorPopi commented on issue mrq/ai-voice-cloning#454 2023-11-27 09:50:34 +00:00
StyleTTS2 - New Free Voice Cloning TTS

To be followed indeed...!

DoctorPopi commented on issue mrq/ai-voice-cloning#445 2023-11-13 13:32:54 +00:00
Can emotion be trained?

Hey! I am in the process of trying to label "intense" emotions with capital letters in the JSON retranscription, as well as with "!!!!" that sort of thing. I've had some pretty interesting results…

DoctorPopi pushed to qol_improvements at DoctorPopi/ai-voice-cloning-fork 2023-11-07 19:50:57 +00:00
8af8e6b6ca quick fic of model path in output name
DoctorPopi opened issue mrq/ai-voice-cloning#441 2023-11-06 06:53:36 +00:00
[Question] Learning Rate Schedule default?
DoctorPopi pushed to qol_improvements at DoctorPopi/ai-voice-cloning-fork 2023-11-04 16:11:18 +00:00
0a09610c2c dlas commit info
DoctorPopi pushed to qol_improvements at DoctorPopi/ai-voice-cloning-fork 2023-11-04 16:08:37 +00:00
3621f30433 added more info on output generated clips
DoctorPopi created branch qol_improvements in DoctorPopi/ai-voice-cloning-fork 2023-11-04 16:04:46 +00:00
DoctorPopi pushed to qol_improvements at DoctorPopi/ai-voice-cloning-fork 2023-11-04 16:04:46 +00:00
DoctorPopi created repository DoctorPopi/ai-voice-cloning-fork 2023-11-04 15:55:33 +00:00
DoctorPopi commented on issue mrq/ai-voice-cloning#435 2023-11-02 18:04:27 +00:00
[Discussion] What is a large dataset?

Oh wow 50k! Yeah I definitely won't be able to produce that many clips. All I have are 4 audiobooks and 600 game character lines. The voice actor had a pretty wide range of voices so that…