README.md · Chamin09/ai_agents

metadata

title: Deep Dive Analysis with Sustainable AI
emoji: 🌿
colorFrom: green
colorTo: blue
sdk: gradio
sdk_version: 5.19.0
app_file: app/main.py
pinned: false
license: mit
tags:
  - sustainability
  - multi-agent
  - nlp
  - computer-vision
  - langchain

Deep Dive Analysis with Sustainable AI A multi-agent AI system for analyzing text and image content on a specific topic, with a focus on sustainability and energy efficiency.

Overview This application allows users to upload text files and images related to a topic, and receive a comprehensive analysis and report. The system uses a combination of AI models for text analysis, image processing, and report generation, all while optimizing for energy efficiency and sustainability.

Key features:

Text analysis with semantic understanding Image captioning and relevance assessment Comprehensive report generation with confidence levels Sustainability metrics tracking Energy-efficient model selection Architecture The system is built with a multi-agent architecture:

Text Analysis Agent: Processes text files to determine relevance and extract key information Image Processing Agent: Captions images and determines their relevance to the topic Report Generation Agent: Creates comprehensive reports based on the analyses Metrics Agent: Tracks sustainability metrics and resource usage Coordinator Agent: Orchestrates the workflow between agents These agents are supported by:

Model managers for text, image, and summarization Utilities for token management, caching, and metrics calculation Communication and synchronization components Installation Clone the repository: git clone https://github.com/yourusername/deep-dive-analysis.git cd deep-dive-analysis Copy Create a virtual environment: python -m venv venv source venv/bin/activate # On Windows: venv\Scripts\activate Copy Install dependencies: pip install -r requirements.txt Copy Usage Running the Application python app/main.py Copy This will start the Gradio web interface, accessible at http://localhost:7860.

Command Line Options python app/main.py --config path/to/config.yaml --log-level INFO --port 7860 --share Copy --config: Path to configuration file (default: config/config.yaml) --log-level: Logging level (default: INFO) --port: Port for the web interface (default: 7860) --share: Create a shareable link Using the Web Interface Enter a topic for deep dive analysis Upload text files related to the topic Upload images related to the topic Click "Start Analysis" View the results in the different tabs: Executive Summary Detailed Report Text Analysis Image Analysis Raw Data Sustainability Features The application includes several features to optimize energy usage:

Token Optimization: Minimizes token usage for LLM operations Adaptive Model Selection: Uses smaller models when appropriate Caching: Avoids redundant computation Smart Routing: Directs tasks to the most efficient components Sustainability Metrics: Tracks energy usage and carbon footprint Configuration The application is configured through config/config.yaml. Key configuration sections include:

app: General application settings token_manager: Token budget and energy coefficients cache_manager: Cache size and TTL settings metrics_calculator: Carbon intensity and PUE values models: Model selection for different tasks agents: Agent-specific parameters Contributing Contributions are welcome! Please feel free to submit a Pull Request.

Fork the repository Create your feature branch (git checkout -b feature/amazing-feature) Commit your changes (git commit -m 'Add some amazing feature') Push to the branch (git push origin feature/amazing-feature) Open a Pull Request License This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments This project uses models from Hugging Face Built with LangChain, PyTorch, and Gradio Inspired by research on energy-efficient AI systems