gradio_article.md

Description

Hey there,

I made this cool app, mainly for my little son. It's a fun voice-controlled thing where he can chat with it and get spoken answers. For example, he can ask it to tell a story or find out stuff he's curious about.

Here's how it works:

I used a bunch of libraries like openai, playsound, gtts, speech_recognition, and gradio. The app records what you say into a microphone and sends it to OpenAI for answers. Then, it turns those answers into speech and plays them back to you.

There's a "transcribe" function that takes audio and a language code, listens to what you say, asks OpenAI for answers, turns those answers into speech, and stores it temporarily. Then it gives you the path to the spoken response.

The app also has a simple user interface made with gradio. You can pick a language and talk into the microphone. The app transcribes your words, gets answers, and plays them back to you.

I'm thinking of adding even more fun stuff like custom voices from VALL-E, Mozilla TTS, OpenSeq2Seq, Vbee, FPT, or Elevenlabs. So I can train my own voice to use for the generated response. Isn't it interesting?

This is a fun and easy way to have voice conversations and learn new things. So, anyone can enjoy it!