Jarvis - Offline Personal Assistant

This is an open-source, fully offline personal assistant inspired by Jarvis. It uses a wake word to activate, transcribes speech to text, processes it with a local LLM, and speaks the response.

Technology Stack

• Wake Word: OpenWakeWord
• Speech-to-Text (STT): Whisper.cpp
• LLM Server: Ollama
• LLM Model: Mistral (7B) new Phi3:mini
• Text-to-Speech (TTS): Piper TTS
• UI Framework: Flask

Setup

1. Create the File Structure:

   • Create all the folders and empty files as laid out in the project structure.

2. Install Python dependencies:

   pip install -r requirements.txt

3. Install/Compile External Tools:

   • Ollama: Download and install from https://ollama.com/. Then run ollama pull mistral or ollama pull phi3:mini.
   • Whisper.cpp: Compile from source from its official GitHub repository.
   • Piper: Download the executable for your OS from the Piper TTS GitHub releases page.

4. Download Audio Models:

   • Download the required models for OpenWakeWord, Whisper.cpp, and Piper TTS.
   • Place them in the corresponding subdirectories inside /audio_models.
   • Important: Update the file paths at the top of stt_module.py and tts_module.py to point to your compiled executables and downloaded models.

5. Run the application:

   python  main.py

   Then, open your web browser to http://127.0.0.1:5000.