This repository will have different projects using AutoGen and Tutorials
JavaScript 0.90%Python 82.44%CSS 1.19%HTML 15.48%
ai's Introduction
Repository - AI Projects/Learning
This repo will be helpful in understanding AutoGen providing examples including prompts and agents for SAAS products, how AutoGen works, and diving into the functionality.
MemGPT has been updated recently and if we don't use memgpt configure to set the openai_key, then it won't work with OpenAI API. I opened issue here: https://github.com/tylerprogramming/ai/issues/1
issues with function calling connected with LM Studio. GPT function calling works, but as soon as the config is swapped for localhost to LM Studio, they are ignored
NEED to make sure that if using LM Studio, set the UserAgent to have a default auto reply to "..." or something. LM Studio complains about this because of the interaction
Projects
autogen_memgpt - understanding integration of MemGPT into AutoGen as an AI Agent
autogen_memgpt_lmstudio - using a local llm to integrate MemGPT into AutoGen with a local server produced by LMStudio
autogentest - examples of basic usage of AutoGen
autogen_functions - learn how to use functions with AutoGen
autogen_multiple_configs - learn how to use multiple configurations in order to use multiple models with AutoGen
autogen_transcribe_video - learn how to use AutoGen and GPT to transcribe and translate a video with functions
ai_agency_01_workout_plan - create your first agency with functions and saving the output to a .txt and .csv file
ai_agency_02_lmstudio - learn how to directly connect to any open-source llm using lm studio with a few examples
autogen_webscraping - use gpt-vision4 with AutoGen to take pictures (web scrape) and use AI to describe the web page
Upcoming Ideas/Projects for Videos
GPT-4 Vision with AutoGen
Video on all available Agent Types AutoGen
AutoGen with CodeInterpreter
AutoGen with TeachableAgent (uses Vector DB to remember conversations)
Auto Generated Agent Chat: Hierarchy flow using select_speaker
Updated MemGPT with new Coding Project
AutoGen Teams, actually creating separate teams that each do a specific thing and pass on what they accomplished to the next one
Combining GPT-4 Vision with a library that can take a screenshot of a website, perhaps with stocks for example, and examine it
Create a Sudoku Puzzle Creator/Checker with an AI WorkForce
Create WebScraper with Puppeteer
Create AutoGen with Whisper
Fitness Tracker with multiple models and LMStudio for LocalLLM