Skip to content

d6e/BabbleBoop

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

45 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

BabbleBoop

This is a fun little project I put together to experiment with real-time translation in VRChat using Large Language Models (LLMs). It's not perfect, but it's been interesting to play around with!

The fun thing about LLMs is that they can translate to anything! English -> Japanese, English -> Emoji, English -> Sophisticated British Butler!

What It Does

  • Listens to your speech in VRChat
  • Transcribes what you say using OpenAI's Whisper
  • Translates the text using GPT models
  • Sends the translation to VRChat's chat box via OSC

Why LLMs?

I wanted to see if using LLMs like GPT could provide more context-aware translations compared to traditional methods. It's not always better, but it can handle some VRChat-specific lingo pretty well!

What You Need

  • A PC running VRChat
  • A microphone
  • VRChat with OSC enabled
  • An OpenAI account (for API access)

Getting Started

Running It

  1. Start VRChat and enable OSC
  2. Populate the config.toml file with your OpenAI API key and target language. Make sure it's in the same folder as the executable. I left an example config file in the repo.
  3. Run the translator:
    • If using a pre-built release, just double-click the executable
    • If you've built from source, use cargo run --release
  4. Start chatting in VRChat!

Some Cool Things It Does

  • Uses a noise gate to ignore background noise
  • Waits for pauses in speech before translating
  • Shows the "typing" indicator in VRChat while it's working
  • Limits API requests to avoid burning through your OpenAI credits too fast

Known Issues

  • Sometimes struggles with very short phrases
  • Can be a bit slow if you talk for a long time without pausing
  • Occasionally makes weird translations (but that can be funny too)

Troubleshooting

If it's not working:

  1. Double-check your VRChat OSC settings
  2. Make sure your mic is working
  3. Check your config.toml file for typos
  4. Verify your OpenAI API key is valid

Want to Tinker?

Feel free to fork the project and make changes! If you come up with any cool improvements, I'd love to see them. If you want to build from source:

  1. Make sure you have Rust installed
  2. Clone this repo
  3. Run cargo build --release

Disclaimer

This is just a personal project and isn't officially associated with VRChat or OpenAI. Use at your own risk!

About

An LLM based translator for VRCHAT OSC

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages