AGI Pipeline

Overview

A comprehensive, modular AGI (Artificial General Intelligence) pipeline integrating state-of-the-art NLP, Computer Vision, and Speech Processing capabilities. This framework is designed to facilitate seamless integration and interaction between different AI modules, enabling the development of sophisticated AI applications.

Features

• Natural Language Processing (NLP): Text generation and summarization using models like T5 and BART.
• Computer Vision (CV): Object detection with YOLOv8 and image classification with ResNet50.
• Speech Processing: Speech-to-text with Whisper (STT) and text-to-speech with Pyttsx3 (TTS).
• Multi-Modal Integration: Understanding scene context by combining text and image inputs.
• Reinforcement Learning (RL): Training agents using PPO in custom environments.
• Real-Time Processing: Handling live video and audio streams for low-latency analysis.

Installation

1. Clone the repository:

   git clone https://github.com/OneFineStarstuff/AGI-Pipeline.git
   cd AGI-Pipeline

2. Set up a virtual environment:

   python3 -m venv venv
   source venv/bin/activate  # On Windows use `venv\Scripts\activate`

3. Install dependencies:

   pip install -r requirements.txt

4. System Dependencies: Ensure ffmpeg and espeak-ng are installed for speech processing.

Usage

Running the API

1. Start the FastAPI application:

   uvicorn main:app --reload

2. Access the Interactive Documentation: Open http://127.0.0.1:8000/docs in your browser to explore the API endpoints.

Using Docker

1. Build the image:

   docker build -t agi-pipeline:1.0.1 .

2. Run the container:

   docker run -p 8000:8000 agi-pipeline:1.0.1

Governance & Compliance

This project enforces strict governance standards for AGI development.

• Governance Artifacts: Located in gstack_artifacts/.
• Validation: Run make verify-governance to ensure all artifacts meet compliance requirements.
• Monitoring: omni_sentinel_24h_monitor.py tracks G-SRI and attestation status.

Contributing

We welcome contributions! Please see our CONTRIBUTING.md for guidelines and CODE_OF_CONDUCT.md for our community standards.

Citation

If you use this software in your research, please cite it as follows:

@software{Tun_AGI-Pipeline_2024,
author = {Tun, Kyaw T.},
doi = {10.5281/zenodo.14504697},
month = {12},
title = {{AGI-Pipeline}},
url = {https://github.com/OneFineStarstuff/AGI-Pipeline},
version = {1.0.0},
year = {2024}
}

Refer to CITATION.cff for more details.

License

This project is licensed under the MIT License.