So is this abandoned? #484

I ran the update and it updated a few modules, but looks like it broke some stuff; transcribe doesn't work half the time, and when it does, it generates an empty file for train.

I haven't seen any updates for six months...is there any hope?

If it is dead, anyone have suggestions on replacements with a UI and good results?

I ran the update and it updated a few modules, but looks like it broke some stuff; transcribe doesn't work half the time, and when it does, it generates an empty file for train. I haven't seen any updates for six months...is there any hope? If it is dead, anyone have suggestions on replacements with a UI and good results?

I suspect it is, yes. Is there a command that can see what versions of software are installed? I have a functional copy of the software on my computer (I haven't trained in awhile...). I could run that command.

I suspect that the creator was hired to work on a commercial product based on his programming here (or was paid to stop development and communication?). Although the result is slow, I'm getting exactly what I have expected from this program. The training is really good.... I haven't look at any of the other free, local TTS products...

I suspect it is, yes. Is there a command that can see what versions of software are installed? I have a functional copy of the software on my computer (I haven't trained in awhile...). I could run that command. I suspect that the creator was hired to work on a commercial product based on his programming here (or was paid to stop development and communication?). Although the result is slow, I'm getting exactly what I have expected from this program. The training is really good.... I haven't look at any of the other free, local TTS products...

That would make sense if that did happen to him. Good for him though, he made a really fun thing, just wish he was able to help out or update if that's the case. And i agree, this has been the best way to clone things for fun.

Anyways, here's what i found out, i'm not a python guru so take this with a grain of salt, or if anyone else can think of a better way/correct me, please do, and please make sure it looks ok, i don't want to damage an install or do anything detrimental.

Activate the virtual environment - navigate to the root directory (for me it was J:/AI/MRQ/ai-voice-cloning), and then type venv\scripts\activate. That should put you in the venv for it. I did this is an admin-level command prompt.
type in pip list --local, that should list all packages installed and the version, and some had editable project locations.

2a. As for exporting...i found online the command pip freeze > XXX.txt, where XXX is the name of the file you want to export that contains versions for the installed packages. Word of warning though, it'll overwrite same name files without a warning, as i named it requirements.txt and it overwrote the one in the root directory.

That would make sense if that did happen to him. Good for him though, he made a really fun thing, just wish he was able to help out or update if that's the case. And i agree, this has been the best way to clone things for fun. Anyways, here's what i found out, i'm not a python guru so take this with a grain of salt, or if anyone else can think of a better way/correct me, please do, and please make sure it looks ok, i don't want to damage an install or do anything detrimental. 1. Activate the virtual environment - navigate to the root directory (for me it was J:/AI/MRQ/ai-voice-cloning), and then type `venv\scripts\activate`. That should put you in the venv for it. I did this is an admin-level command prompt. 2. type in `pip list --local`, that should list all packages installed and the version, and some had editable project locations. 2a. As for exporting...i found online the command `pip freeze > XXX.txt`, where XXX is the name of the file you want to export that contains versions for the installed packages. Word of warning though, it'll overwrite same name files without a warning, as i named it requirements.txt and it overwrote the one in the root directory.

Here is my requirements.txt:

git+https://github.com/openai/whisper.git

more-itertools
ffmpeg-python
gradio<=3.23.0
music-tag
voicefixer
psutil
phonemizer
pydantic==1.10.11
websockets

Here's the pip -- list local with just that software:

openai-whisper 20230314 [my comment, this is probably when I built the software program]
more-itertools 10.1.0
ffmpeg-python 0.2.0
gradio 3.23.0
music-tag 0.4.3
voicefixer 0.1.2
psutil 5.9.5
phonemizer 3.2.1
pydantic 1.10.11
websockets 11.0.3

I suspect the issues you are having has gotta be with the open-ai whisper install.

Here is my requirements.txt: git+https://github.com/openai/whisper.git more-itertools ffmpeg-python gradio<=3.23.0 music-tag voicefixer psutil phonemizer pydantic==1.10.11 websockets Here's the pip -- list local with just that software: openai-whisper 20230314 [my comment, this is probably when I built the software program] more-itertools 10.1.0 ffmpeg-python 0.2.0 gradio 3.23.0 music-tag 0.4.3 voicefixer 0.1.2 psutil 5.9.5 phonemizer 3.2.1 pydantic 1.10.11 websockets 11.0.3 I suspect the issues you are having has gotta be with the open-ai whisper install.

One more thing, I also have a few programs installed directly:

TorToiSe 2.4.5 D:\APPLICATIONS\ai-voice-cloning-2\ai-voice-cloning\modules\tortoise-tts
dlas 0.0.1 d:\applications\ai-voice-cloning-2\ai-voice-cloning\modules\dlas

One more thing, I also have a few programs installed directly: TorToiSe 2.4.5 D:\APPLICATIONS\ai-voice-cloning-2\ai-voice-cloning\modules\tortoise-tts dlas 0.0.1 d:\applications\ai-voice-cloning-2\ai-voice-cloning\modules\dlas

Looking through my install notes.... I also had to do this at one point:

pip uninstall whisper
pip install openai-whisper

Looking through my install notes.... I also had to do this at one point: pip uninstall whisper pip install openai-whisper

Awesome, thank you for posting the versions...the weird thing is, i tried using what you listed and didn't have any luck, but on a whim i tried an old install i had and it started working...i'm honestly confused as to why it started working again as i ran the update on that one; hence why i tried starting fresh. Training with defaults creates both training and validation files, and i was able to train and use a model last night...so not sure what happened.

But it's good to have some versioning that's confirmed working just in case something happens like this again.

Awesome, thank you for posting the versions...the weird thing is, i tried using what you listed and didn't have any luck, but on a whim i tried an old install i had and it started working...i'm honestly confused as to why it started working again as i ran the update on that one; hence why i tried starting fresh. Training with defaults creates both training and validation files, and i was able to train and use a model last night...so not sure what happened. But it's good to have some versioning that's confirmed working just in case something happens like this again.

Any chance you could document what commit ID your install is using? I'm having issues setting up the app as well

So is this abandoned?

Pretty much.

The extremely short of it, is that this repo itself is in dire need of a rewrite.

If it is dead, anyone have suggestions on replacements with a UI and good results?

https://github.com/JarodMica/ai-voice-cloning caught my eye some weeks (or months) ago. I haven't tested it myself, but I'll give it my blessing since it's receiving the love I cannot give it.

I suspect the issues you are having has gotta be with the open-ai whisper install.

Sounds right. The whispers included in the requirements and if you elect to use WhisperX will break dependencies and require a different set of pip installs.

> So is this abandoned? Pretty much. The *extremely short* of it, is that this repo itself is in dire need of a rewrite. > If it is dead, anyone have suggestions on replacements with a UI and good results? https://github.com/JarodMica/ai-voice-cloning caught my eye some weeks (or months) ago. I haven't tested it myself, but I'll give it my blessing since it's receiving the love I cannot give it. --- > I suspect the issues you are having has gotta be with the open-ai whisper install. Sounds right. The whispers included in the requirements and if you elect to use WhisperX will break dependencies and require a different set of `pip install`s.

@mrq I just wanna say thank you for even putting this thing out, without it i wouldn't have gone down the rabbit hole of all this TTS stuff.

I actually started using that one, as it looks like he's been basing his stuff off of your work, plus he put out a bunch of youtube videos too to help the community.

Again though, thanks for all of your hard work, i totally understand that it's a multitude of work and sometimes life doesn't want to be nice to your hobbies.

@mrq I just wanna say thank you for even putting this thing out, without it i wouldn't have gone down the rabbit hole of all this TTS stuff. I actually started using that one, as it looks like he's been basing his stuff off of your work, plus he put out a bunch of youtube videos too to help the community. Again though, thanks for all of your hard work, i totally understand that it's a multitude of work and sometimes life doesn't want to be nice to your hobbies.

This repo is indeed the best voice cloning tool I've tested for nearly a year, nothing except Eleven Labs gets better results than this...(I would even say that it surpasses Eleven Labs in some aspects such as retaining the accents)

Thank you @mrq for all the hard work and for pointing to JarodMica's repo :)

This repo is indeed the best voice cloning tool I've tested for nearly a year, nothing except Eleven Labs gets better results than this...(I would even say that it surpasses Eleven Labs in some aspects such as retaining the accents) Thank you @mrq for all the hard work and for pointing to JarodMica's repo :)

Labels Milestones

So is this abandoned? #484