[Enhancement] Pinokio Script

[emrahtoy]

I have created a https://pinokio.computer script and released here.

[Ionaut]

I used this and it works well. needed to run update after first install to apply some repo changes though!

[emrahtoy]

I used this and it works well. needed to run update after first install to apply some repo changes though!

Thank you, I will check and apply changes if needed.

[JarodMica]

Appreciate the contribution! I will leave this open in case more people would like to try it out - I cannot verify or validate that the script works as I haven't played around with Pinokio

[theslipperyCarrot]

I installed the StyleTTS via Pinokio with success. I´m very interested in this concept!
Till now I´ve installed 12 different ai-WebUis. Every time installing cuda, torch, hubertbase etc. in a venv... every time errors and conflicts. Its not easy! I hope Pinokio will help!

Will StyleTTS installed via Pinokio work with Audiobookmaker V3? Otherwise it would not be that interesting for me.
Thanks and best regards!

[JarodMica]

Will StyleTTS installed via Pinokio work with Audiobookmaker V3? Otherwise it would not be that interesting for me. Thanks and best regards!

You can train models in the pinokio version of it, which will be needed for the AM v3. The Audiobookmaker V3 will have StyleTTS built into it as an installable engine, so venv/conda will need to be used. The headache of managing all that will be handled in the packaged verions I distribute to members, but I'm afraid if you are going to be using it through manual installation, you may run into the same issues. I'm trying going to try and make it as easy as possible, but issues may still crop up

[theslipperyCarrot]

I'm eagerly awaiting the new Audiobookmaker and of course I'll do my best to install it. I had also successfully installed StyleTTS and was able to use it once. The next day, unfortunately, I got this RuntimeError: #8
Should I open a new issue, because this one is closed?

Thanks to your work @JarodMika!
Thanks to your work @emrahtoy!
I would appreciate if Pinokio would become a new standard for ai-UIs for a easy Installation without all the conflicts and the double-installations.

[emrahtoy]

Hello @theslipperyCarrot
This system has a dependency on external packages/apps which is why you are facing the problem. To solve this issue you must follow the steps below;

1. Install espeak-ng
2. Normally, it changes system path environments, but if it does not change, here are 2 approaches to fix it. First, add the espeak-ng path into the path variable in the system environment variables ( for windows ) or use the customize option in the pinokio system. The Pinoiko system is the easiest way for me.

Here are screenshots from a fresh install with pinokio;

Click on the "Customize" button. This button is everywhere so don't be alarmed if you can't see it in the same way.

Then you must set 2 variables;

1. PHONEMIZER_ESPEAK_PATH
2. PHONEMIZER_ESPEAK_LIBRARY

[theslipperyCarrot]

Hello!
Thank you very much for your answer!
Unfortunately, I expressed myself in a misleading way.
I installed StyleTTS twice, once via Pinokio, but also once in the classic way via CMD. The installation via Pinokio worked well and runs flawlessly! I had only done the installation via CMD additionally, because I had thought this would be necessary for the Audiobookmaker V3. But that's not the case.
I had installed espeak-ng and added it to path. Perhaps the two installations interfere with each other?

I will now simply do the training via the instance at Pinokio and wait for the new version of the Audiobookmaker.
I think Pinokio's concept is great and hope that there will be more about it in the future!
Thank you for your work and best regards!