AllTalk TTS Update

[erew123]

I didn't know about your project, someone had a problem so I've made a small update for you:

Here's a summary of the changes for the Pull Request:

• Version Support

  • Added support for both AllTalk V1 and V2 servers
  • Automatic URL handling for different version responses (V1: full URL, V2: relative path)

• Voice Selection

  • Standard voice selection dropdown with server-fetched voices
  • Voice preview functionality
  • Persistent voice settings stored in local storage

• RVC Integration (AllTalk V2)

  • RVC voice selection with fetched available voices
  • RVC pitch control (-24 to +24)
  • RVC settings automatically disabled for V1 servers
  • Defaults to 'Disabled' for RVC voice and '0' for pitch

• Server Configuration

  • Server URL configuration with validation
  • Server status monitoring
  • Automatic reconnection and voice list refresh
  • Version selection (V1/V2)

• UI Improvements

  • Clean, consistent interface matching existing design
  • Language selection dropdown
  • Status indicators for server connection
  • Loading states and error handling
  • Documentation link to AllTalk Wiki

• Settings Persistence

  • Settings saved in local storage
  • Settings restored on page reload
  • Independent settings for V1/V2 configurations

• Error Handling

  • Comprehensive error reporting
  • Server connection validation
  • Fallback handling for unavailable features

Thanks

[vercel]

The latest updates on your projects. Learn more about Vercel for Git ↗︎

Name	Status	Preview	Comments	Updated (UTC)
amica	✅ Ready (Inspect)	Visit Preview	💬 Add feedback	Nov 28, 2024 5:28pm

[vercel]

@erew123 is attempting to deploy a commit to the heyamica Team on Vercel.

A member of the Team first needs to authorize it.

[erew123]

Oh, I forgot to mention. This now plays audio in the web-browser and no longer using the AllTalk terminal/command prompt, so that should be better generally.

[slowsynapse]

Thank you, I was away on business. Much appreciated, will review. :))

[slowsynapse]

Currently, can't get it to run on Mac OS X, some dependency issue ("TTS" and "sudachipy"), trying to diagnose.

[erew123]

Hi @slowsynapse Sorry its taken a while to get back to you, Im mid releasing AllTalk v2 the other day and its crazy busy. Do you mean youre installing AllTalk on a mac? or the extension is having an issue on mac?

With AllTalk on mac, Im a bit shaky on that as Ive never had a mac to test it and Ive put up the gauntlet to someone to get it working and Ill try help as/where I can. If it is specially just getting a copy of AllTalk for testing though, you can use a free Google Colab.

My Wiki Page on it is here https://github.com/erew123/alltalk_tts/wiki/Google-COLAB

But you can go to this link: https://colab.research.google.com/github/erew123/alltalk_tts/blob/alltalkbeta/googlecolab.ipynb and sign into your google account.

Then it should automatically set a T4 server type (top right) and:

and once you are past Step 2 you should get this:

Its the API address you would want with Amica, though it will only have set up 1x voice at this point in time, on Piper TTS, to download other voices/TTS models etc you would have to go into the AllTalk Gradio interface (the links below the API) and get to know AllTalk a bit. If you want to go that far, you can either look around the help in the interface or look on the AllTalk Wiki quick start. https://github.com/erew123/alltalk_tts/wiki/AllTalk-V2-QuickStart-Guide

Give me a shout if you need.

Thanks

[slowsynapse]

hey @erew123 , we identified the issue and its a slight syntax issue, will be sending update soon. We got it working too, it's super cool.

[slowsynapse]

^ Here is what needs fixing (in comments), Line 25 and Line 81

[erew123]

@slowsynapse Ill take that as a win (just the 2x lines wrong). So busy with everything else and I knocked this together kind of quick (if Im honest). Should be updated as you specified. Thanks for debugging, saved me a hell of a task there!!