This project provides a local, privacy-focused voice chatbot that integrates Ollama for language processing, Hugging Face Transformers for speech recognition, and Coqui TTS Toolkit for speech synthesis. It's designed for users who want an offline conversational AI experience, supporting various interaction modes from text-only to full voice input/output.

How It Works

The chatbot orchestrates three core AI components: Ollama for LLM inference, Hugging Face's Whisper for speech-to-text (STT), and Coqui TTS for text-to-speech (TTS). Users can interact via text or voice. Voice input is automatically detected without a wake word, and a short period of silence signals the end of speech. The system is highly configurable, allowing users to swap models, specify devices (CPU/GPU), and disable components like STT or TTS for different interaction styles.

Quick Start & Requirements

• Install: pip install git+https://github.com/mezbaul-h/june.git@master or clone and pip install .
• Prerequisites: Ollama, Python 3.10+, PortAudio development package (Linux/macOS), Microsoft Visual C++ 14.0+ (Windows).
• Usage: Pull an Ollama model (e.g., ollama pull llama3.1:8b-instruct-q4_0) and run june-va. Configuration can be customized via a JSON file.
• Docs: Customization Section

Highlighted Details

• Supports multiple interaction modes: Text-only, Voice-to-Text, Text-to-Audio, and Voice-to-Audio.
• Automatic voice activity detection without a wake word.
• Configurable via JSON for model selection, device allocation (CPU/GPU/MPS), and disabling components.
• Supports voice cloning with compatible Coqui TTS models (e.g., xtts_v2).
• Can connect to remote Ollama instances via OLLAMA_HOST environment variable.

Maintenance & Community

No specific community links (Discord/Slack) or roadmap are provided in the README. The project appears to be maintained by a single author, mezbaul-h.

Licensing & Compatibility

The README does not explicitly state a license. This requires further investigation before commercial use or integration into closed-source projects.

Limitations & Caveats

The project does not specify a license, which is a significant caveat for adoption. The README also lacks details on performance benchmarks or specific hardware requirements beyond general prerequisites.