You can add this plugin by one of:
Change the plugin preferences to something like https://tts.yourdomain.com/api_v2 or whatever your actual domain is and make sure you have your sovits api_v2.py
api running on that url.
If you want some of the API introspection quality-of-life improvements (listbox of available voice samples, etc) use my patched api_v2.py at https://github.com/cpumaxx/GPT-SoVITS/blob/main/api_v2.py
You may need to use the -c parameter to point to a custon YAML file that defines things like your preferred ckpt and pth files.
You need at least one character+emotion defined to use this plugin.
You will need to put any sound samples you want to use on your SoVITS server, in the same folder as the api_v2.py script.
Be very careful to use a high-quality voice samples between 3-10 seconds and set the "Prompt Text" to precisely what is being said in the sample.
You can set up SoVITS and get the required files from these resources:
https://github.com/RVC-Boss/GPT-SoVITS/
https://rentry.org/GPT-SoVITS-guide
https://huggingface.co/cpumaxx/SoVITS-anime-mini-tts
git clone https://github.com/cpumaxx/sovits-ff-plugin
<all_urls>
to your API endpoint and change the ID (you can't re-use them since they will be rejected my Mozilla signing)zip
to xpi