LLM based analysis

Author: Omer Ansari

Initialize date: 5.27.24

Summary

This repository provides the code for a local environment for AI-driven development. It includes code for doing AI development both using an IDE (vscode), or a jupyter notebook.

It provides a local environment setup (MacOS only) that can be

• used with OpenAI's APIs (link),
• opensource LLMs hosted on the laptop (link), and
• a small scrum team (using multi-agent framework backed by local LLMS) (link).

In addition, I'll be adding tools for

steps to set up the jupyter notebooks locally in a virtual environment:

• git clone git@github.com:knail2/llm-based-analysis.git
• cd llm-based-analysis/
• python -m venv venv
• . ./venv/bin/activate
• pip install -r requirements.txt
• pip install --upgrade 'jupyter-server<2.0.0' # horked library called tornado breaks notebook, so downgrade jupyter
• jupyter contrib nbextension install --user #set up jupyter extensions
  • the article is here
  • the documentation for the extensions here
• jt -t onedork -fs 95 -altp -tfs 11 -nfs 115 -cellw 88% -T # optional, dark mode
• jt -r # optional, remove dark mode
• jupyter notebook # run notebook