Overview

Jamboxx Infinite is a Unity-based application featuring:

• Multiplayer functionality via Unity Networking
• Offline AI chatbot powered by Mistral-7B using llama.cpp
• AI Voice Conversion (Not included in the first release. You may activate the relevant Unity page and refer to the jamboxx-infinite-backends repo for more info.)

Repository Structure

organisation: jambox-infinite

Subdirectories:

• Jamboxx_infinite – Unity application
• website – Website frontend
• Jamboxx-infinite-backend-llama – LLaMA chatbot backend

System Requirements

• Unity (please specify version)
• Git
• CMake (for building llama.cpp backend)
• Compatible GPU (if running quantized Mistral model with GPU acceleration)
• Windows / macOS / Linux (mention supported OS for builds)

Deployment Steps

1. Clone the Main Repository

git clone https://github.com/jamboxx-infinite/Jamboxx_Infinite.git
cd Jambox_infinite

2. Build the Unity App

• Open Unity Hub
• Add the project located in Jamboxx_infinite/
• Select target platform (e.g., Windows)
• Build the application via File > Build Settings

3. Set Up the LLaMA Chatbot Backend

Option A: Use Precompiled Version

Download the precompiled llama backend binary and place your model files in the correct directory.

Option B: Build Manually

git clone https://github.com/jamboxx-infinite/Jamboxx-infinite-backend-llama.git
cd jamboxx-infinite-backend-llama
mkdir build && cd build
cmake ..
make

• Download the Mistral-7B model or the model you wish to use (GGUF format) and place it in the expected models/ directory.
• Launch the backend:

./main --model models/mistral-7b-instruct-v0.2.Q4_K_M.gguf --port 8001

(Adjust flags as needed; refer to README.md inside the backend repo for detailed options)

4. Connect Unity App to Backend

• Ensure the Unity app is configured to connect to localhost:8001 (or your backend server IP)
• Launch the Unity app

Notes

• AI Voice Conversion is disabled by default in the first publish. To enable it:
  • Activate the voice conversion page in the Unity Editor
  • Refer to the Voice Backend Docs for setup instructions

Troubleshooting

• Unity crashes or model not found: Check that model path in llama backend matches the filename and directory
• Chatbot not responding: Ensure llama backend is running, and port/IP is correct
• Multiplayer issues: Check Unity Network configuration and firewall settings

Credits & Links

• Main App Repo: Jambox-infinite
• LLaMA Backend: jamboxx-infinite-backends-llama
• Unity Hub: https://unity.com/download