Skip to content

BeeBrain is your personal chatbot. Use tools, generate images, run code and so much more!

License

Notifications You must be signed in to change notification settings

MartianInGreen/BeeBrain

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

46 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Overview of BeeBrain

BeeBrain is an advanced, multi-functional chatbot designed for users who seek a comprehensive digital assistant. It combines state-of-the-art machine learning models with a range of interactive tools to provide a versatile and user-friendly experience. Whether you're looking to run complex code, search the web efficiently, generate images, or simply engage in a conversation, BeeBrain is equipped to meet a wide array of needs.

Key Features:

  • Diverse Toolset: Includes a web browser, image generator, text/PDF analyzer, file browser, Jupyter Notebook, and more.
  • Advanced Chat Interface: Utilizes modern Large Language Models (LLMs) for seamless interaction.
  • Customizability: Designed for users who enjoy tweaking and extending functionality through Python coding.
  • Future-Ready: Plans for integrating Vision with GPT-4-Turbo and other cutting-edge features.

General Note

BeeBrain is primarily targeted at users with a technical background, particularly those comfortable with Python programming. It's not an 'out-of-the-box' solution for end-users but a framework for building and customizing your own chatbot.

Chat Interface Capabilities

BeeBrain's core functionality is accessible through an intuitive LLM chat interface, supporting various models and features.

Supported Models

With Function Calling

  • GPT-4-Turbo (with and without Vision): The latest model with enhanced capabilities, including vision (coming soon).
  • GPT-3.5-Turbo-1106: A robust model offering a balance of performance and efficiency.

Without Function Calling

  • Claude 2.1, Claude 2, Claude Instant v1, PPLX 70B Online, Nous-hermes-llama2-70b, Phind-codellama-34b, Zephyr-7b-beta, Toppy-m-7b: These models offer diverse functionalities but lack tool integration.

Vision Integration

Using the Vision tool all tool calling LLM's are able to see.

Available Tools

BeeBrain comes with an assortment of tools, each designed to enhance the user experience in unique ways.

Current Tools

  • Browser: Enables web searches and website scraping.
  • Image Generation: Powered by DALLE-3 or SD-XL for high-quality image creation.
  • Vision: Enables all models to see using GPT-4-Vison and Llava-13b
  • Jupyter Notebook: Facilitates running and testing Python code.
  • File Browser: LLM can see which files have been uploaded and been created.

Tools in Development

  • Text/PDF Analysis & Summary: For efficient document handling.
  • Paper Search: Assists in academic research.
  • Wolfram|Alpha: Provides expert knowledge and computations.
  • Audio Generation: For creating voiceovers and sound effects.

Roadmap

This roadmap isn't an absolute truth, more some guidelines and goals.

Version 0.1.0

  • Chatting with different LLMs
  • Chat history
  • Tool support
  • Tools: Vision, Image Generation, Python, Web Browser
  • Mission Support (Characters, Specialists, etc.)
  • Basic Ui (Copying, Image Copying, Input & Output)
  • Build ready (Windows)

Version 0.2.0

  • Tools: Text/PDF, Wolfram Alpha, Audio Generation, Video Generation
  • Extended Interface: Seperate Tabs for Image Generation/Editing & Search
  • Extended Interface: Chat Management, Folders, Changing names, etc.
  • Extended Interface: Regenerate, Chat-Branching, Stop Generation, Better handling of adding messages to chat
  • Extended Interface: Prompt Library Support
  • Memory Handling: Auto, 1+Last Prompts, Summarie
  • Better Missions: Individual default Tools
  • Automatic Tools (automatically choose tools)
  • Voice Conversations
  • Better Web Search with Seperate LLM for Search Tasks and focuses (like with Vision)(Focuses: Web, Wikipedia, Academia, YouTube, Reddit)
  • More Platforms: Phone UI & Android App

Version 0.3.0

  • Tools: Doc Maker (PDF, CSVs, Presentations), CodeRunner (C/C++, Go, Rust, Java...), Diagram maker
  • Easy Tools: Add and modify tools within the app
  • Memory Handling: Long Term Memmory
  • More Platforms: Web App, Linux App, Chrome Extention
  • ... more to come

System Requirements

To utilize BeeBrain effectively, users need to obtain API keys from various service providers.

Required API Keys

  • openai: For LLM and image generation.
  • openrouter: LLM routing.
  • replicate: Various tools
  • stabilityai: Image processing.
  • brave-search: Web search functionality.
  • wolfram-alpha: Access to Wolfram Alpha's computational engine.

Images

image image image