Skip to content

Latest commit

 

History

History
137 lines (101 loc) · 3.21 KB

README.md

File metadata and controls

137 lines (101 loc) · 3.21 KB

AI-Companion

Your AI companion to support you during your lows and highs, providing intelligent, real-time assistance.

Overview

The AI Companion is a cutting-edge, low-latency voice assistant that:

  • Converts voice input to text using OpenAI's Whisper.
  • Processes text using Hugging Face's Large Language Model (LLM).
  • Converts responses back to speech with Edge-TTS.
  • Features a web interface and real-time performance optimization for seamless interaction.

Features

  • Voice Activity Detection (VAD): Detects speech activity and ignores silence.
  • Speech-to-Text (STT): High-accuracy transcription using the faster-whisper model.
  • Text-to-Speech (TTS): Converts text to natural speech with tunable pitch, speed, and voice type.
  • LLM Integration: Generates intelligent responses with Hugging Face models.
  • Streamlit Interface: Offers a user-friendly web-based interaction mode.

Installation

Install Daytona

  1. Run the following command to install Daytona:

    curl -sf -L https://download.daytona.io/daytona/install.sh | sudo bash

  2. Start the Daytona server:

    daytona server -y && daytona

  3. If the server daemon stops at any point, restart it using:

    daytona serve

  4. Create a workspace for the AI Companion:

    daytona create https://github.com/Ankur2606/AI-Companion.git --devcontainer-path=.devcontainer/devcontainer.json

Manual Setup

Prerequisites

  • Python 3.8+
  • Git
  • Hugging Face API Token

Steps

  1. Clone the Repository:

    git clone https://github.com/Ankur2606/AI-Companion.git
cd AI-Companion

  2. Set Up a Virtual Environment:

    python -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

  3. Install Dependencies:

    pip install --upgrade pip
pip install -r requirements.txt

  4. Configure Hugging Face Token:

    • Create an account at Hugging Face and generate an API token.
    • Set the token as an environment variable: 
      export HUGGINGFACE_API_TOKEN=your_token  # On Windows: set HUGGINGFACE_API_TOKEN=your_token

Usage

Daytona Mode

Run the project through Daytona for optimized performance:

  1. Start Daytona if not already running:

    daytona serve

  2. Launch the AI Companion:

    daytona run

Manual Mode

  1. Web Interface: Launch the Streamlit web app: 
    streamlit run main.py

Troubleshooting

  • Daytona Server Issues:
    If the server stops, restart it with:

    daytona serve

  • Initial Model Downloads:
    The faster-whisper model may take time to download (~2GB). Ensure a stable internet connection.

  • Environment Variable Issues:
    Verify the Hugging Face API token is correctly set in your environment.

Contributing

We welcome contributions! Please feel free to:

  • Open issues for bugs or suggestions.
  • Submit pull requests for improvements.

License

This project is licensed under the MIT License.