Learning Large Language Model (LLM) frameworks and API that run locally

The primary objective of this repo is to explore setting up Llama 2 to run locally and LLM development frameworks and libraries to provide a foundational runtime environment that can run on on Laptop for further more advance development.

Learning Steps and Tutorials

This repository is structured as a series of progressive learning steps and tutorials. Each step is designed to provide a simple approach of getting started or baseline development environment setup for further development with LLMs running locally. Below is an outline of the learning path:

Step 1: Running LLama 2 locally with LangChain

• Overview: Get Meta Llama 2, quantize it, and setup to run LLM type of hello world with LangChain.
• Code Repo: README.md
• Resources:
  • Meta LLama 2
  • Quantize Llama models with GGML and llama.cpp
  • LangChain

Step 2: Simple Web Chat Interface with Gradio and LangChain

• Overview: Building simple web LLM chat interface interact with LLM running locally.
• Code Repo: README.md
• Resources:
  • gradio - Interface is Gradio's main high-level class, and allows you to create a web-based GUI / demo around a machine learning model (or any Python function) in a few lines of code.

Step 3: LLama / Gradio / RAG

• Overview: Building simple web LLM chat interface interact with LLM and RAG (Retrieval Augmented Generation) running locally. RAG finds the most relevant content or document you specified, and use those extra context for LLM to answer the initial query.
• Code Repo: README.md
• Resources:
  • What is Retrieval-Augmented Generation
  • What we need for Retreival in RAG
  • Control Your Model's Knowledge and Hallucinations