GithubHelp home page GithubHelp logo

sikkha / ai-blue Goto Github PK

View Code? Open in Web Editor NEW
1.0 1.0 0.0 31.83 MB

AI Blue is a multihead AI integration platform for complex analysis. It combines diverse AI insights using advanced routing and weighting techniques, focusing on dynamic, real-world scenarios for nuanced, intelligent responses.

License: Apache License 2.0

Python 100.00%

ai-blue's Introduction

AI Blue

Project Overview

AI Blue is a cutting-edge, multi-headed AI integration platform designed to analyze and respond to complex scenarios with precision. By leveraging a combination of specialized AI models through advanced routing and weighting algorithms, AI Blue delivers nuanced and intelligent insights into a wide array of dynamic real-world problems. This platform stands at the forefront of technology, offering adaptive, informed, and contextually aware solutions.

AI Blue Logo

Getting Started

Prerequisites

Ensure you have Python installed on your system. AI Blue is built to be simple and straightforward, requiring only three main files to function:

  1. requirements.txt - Lists all the necessary Python packages.
  2. ai_blue_prototype.py - The main Python script for the AI Blue model.
  3. session_manager.py - The user session management engine.

Option

We've created a showcase for LLamaindex and Trulen to evaluate RAG and ChromaDB. However, it hasn't been used in AI Blue yet. It will be explored further to generate more documents with the RAG Agent later.

  1. showcase_leaderboard.py - Showcase for LLamaIndex, RAG, ChromaDB and Trulen evaluation leaderboard.

Installation

  1. Clone the repository:
    git clone https://github.com/yourusername/ai-blue.git
  2. Navigate to the AI-blue directory:
    cd ai-blue
  3. Install the required packages:
    pip install -r requirements.txt

Running the Application

To run AI Blue, use the following command:

nohup python session_manager.py &
nohup python ai_blue_prototype.py &

Alternatively, you can use screen for session management, and start the user_management module.

screen
python session_manager.py

And press Ctrl-a, d to exit that session. And start AI Blue module.

screen
python ai_blue_prototype.py

And press Ctrl-a, d to exit that session once again.

You can also use screen commmand screen -ls to monitor and screen -r SCREEN-IDreturn to specific screen session.

AI Blue listens on port 192.0.0.1:7860 and the database session listens to port 192.0.0.1:5001. You can experiment with the live model at:

  • Local: http://192.0.0.1:7860/
  • Public IP: http://34.121.86.8:7860/
  • Short URL: bit.ly/ai-blue

Data Collection and Privacy

AI Blue is designed to continually improve and personalize the user experience. To achieve this, it collects and analyzes session information, including data from cookies and chat logs. This information helps us understand how users interact with AI Blue, allowing us to make enhancements and provide a more tailored experience.

What We Collect:

  • Cookies: To understand user preferences and provide a personalized experience.
  • Chat Logs: To analyze interactions and improve response accuracy and relevance.

Your Privacy: Your privacy is important to us. All collected data is used solely for the purpose of enhancing AI Blue's performance and user experience. We are committed to maintaining transparency and control over your data.

Prefer Not to Share?: If you prefer not to have your data collected, you have the option to use AI Blue in a more private setting:

  • Download and Compile: You can download AI Blue and compile it on your own local system. This way, you can use the AI without sending data back to our servers. The database will be stored in session.db. You can access the database by using the command sqlite3 sessions.db. Within it, you can retrieve data using the command SELECT * FROM sessions;.

By using AI Blue, you agree to the data collection and use as described above. We appreciate your trust and are committed to using your information responsibly and improving your experience with AI Blue.

Concept and Algorithm

AI Blue is a sophisticated chatbot powered by a multi-headed AI approach, designed by AI Tensibility of the Siam Intelligence Unit. This innovative structure allows different AI models to analyze input data concurrently, with each model focusing on its area of expertise. The outputs from these models are then routed through a central "router" model, which intelligently integrates these insights based on predefined weights and contextual understanding. This dynamic system ensures that AI Blue's responses are not only comprehensive but also reflect a deeper level of analysis and insight.

Enhanced by Retrieval-Augmented Generation (RAG), AI Blue can access and incorporate the most current information into discussions, ensuring its knowledge base is always up-to-date. It's not just about processing data; it's about understanding, learning, and adapting to provide the most accurate and relevant information possible.

Key Features:

  • Multi-headed AI Structure: AI Blue's architecture allows for efficient information processing and response generation, enabling it to tackle a wide array of topics and questions with depth and nuance.
  • Enhanced RAG (Retrieval-Augmented Generation): This mechanism empowers AI Blue to seamlessly access and integrate up-to-date information, ensuring that conversations are informed by the latest data and developments.
  • Dynamic Weighting: AI Blue adjusts the influence of each sub-model in real-time based on the context, ensuring that the most relevant expertise is always at the forefront.
  • Intelligent Routing: The central router model analyzes and directs tasks to the most suitable AI model, optimizing the chatbot's understanding and response accuracy.
  • Robust Integration: AI Blue synthesizes outputs from various AI models into a cohesive and coherent response, providing users with comprehensive answers.
  • Dynamic Learning: Continuously learning from interactions, AI Blue refines its accuracy and effectiveness over time, adapting to new information and user feedback.
  • Conversational Adaptability: Tailoring responses to individual communication styles, AI Blue ensures a personalized and engaging interaction experience.
  • Privacy and Security: Adhering to strict security measures, AI Blue treats all conversations confidentially, respecting user privacy and data protection.

AI Blue represents the cutting edge in chatbot technology, offering a conversational experience that's not just informative but also adaptive, engaging, and ever-evolving. Whether you're looking for detailed explanations, current event discussions, or just a friendly chat, AI Blue is equipped to provide a superior interaction experience.

Contributions

We welcome contributions and suggestions! Please read CONTRIBUTING.md for details on our code of conduct and the process for submitting pull requests.

License

This project is licensed under the Apache License 2.0 - see the LICENSE file for details.

Acknowledgments

  • Thanks to all the contributors who have invested their time into making AI-blue a valuable tool.
  • Special thanks to the open-source community for providing the tools and libraries that make projects like this possible.

ai-blue's People

Contributors

sikkha avatar

Stargazers

Chanelle Henry avatar

Watchers

 avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.