GLadOS Voice Assistant is a terminal-based assistant that utilizes OpenAI's API as its language model backbone and uses a PyTorch voice model trained on Ellen McClain's Portal 2 voice lines. It allows you to interact with GLaDOS through chat conversations and hear responses in her distinctive voice.
The core features of the GLadOS Voice Assistant include:
- Chatting with GLaDOS using OpenAI's language model API.
- Generating GLaDOS-like voice responses using voice model.
- Running the assistant directly from the terminal.
- Clone the repository:
git clone https://github.com/Hexanol777/GLadOS-Voice-Assistant.git
- Install the dependencies:
pip install -r requirements.txt
- Install the eSpeak synthesizer:
Follow the installation instructions for eSpeak on your operating system.
- Configure the Assistant:
Edit the config.json file to specify your OpenAI API key, the text generation model you want to use, and the initial prompt.
- Run the Assistant:
- Double-click
glados.bat(Windows) or executeGLadOS.pyin your terminal.
Future enhancements may include:
- Implementation of an offline language model with low resource usage for offline interactions.
- Adding a speech-to-text transcriber to enable voice-based interactions.
- Enabling the assistant to open/start applications upon the user's request.
- May include a GUI in the future
This project is inspired by and utilizes the glados-tts voice generator developed by R2D2FISH.
This project is licensed under the MIT License.
Feel free to contribute, provide feedback, or suggest improvements!