search

BoltzmannEntropy / xtts2-ui

A User Interface for XTTS-2 Text-Based Voice Cloning using only 10 seconds of speech
MIT License
259 stars 42 forks source link

Changed Streamlit UI to Gradio #1

Closed pbanuru closed 11 months ago

pbanuru commented 11 months ago

Hi, I changed the UI to be Gradio-based in order to integrate features such as inbuilt audio-recording, file uploads, and saving target voice recordings. I'm not familiar with Streamlit features, but I thought this may be beneficial, as many popular AI webuis already use Gradio for the interface. Ex. Automatic1111 (Stable Diffusion web-ui, text-generation-webui, and tts-generation-webui

image

I removed the setup files in this version, though it would be good to add them back it.

In addition, I ran into a bug while using the Multilanguage function. I tried using Japanese, and entered Japanese text in the "to convert" box. I got this error: ModuleNotFoundError: No module named 'cutlet' After pip installing cutlet, other bugs appeared. I did not investigate further and for the time being, mentioned in this webui that only English functionality is currently working. I did not make a new issue.

BoltzmannEntropy commented 11 months ago

Thank you so much! I run your version on my OSX and it seems to be working well so I accepted your commits.