README.md

Learning-LLM-From-Zero

This repository is all about my LLM journey: how to use, fine-tuning, and build inference services with LLM.

It's challenging but interesting .

中文版

Table of Contents

• Chapter 1: Run LLAMA2 model with Meta official python code
• Chapter 2: Run LLAMA2 model with HuggingFace transformers
• Chapter 3: Embedding your PDF and send it to LLM.
• Chapter 4: Store moer embedding data with vector database Qdrant
• Chapter 5 Use vLLM build a inference service like openai chatGPT
• Chapter 6: Fine tuning your own model use autotrain-advanced
• Reference

Introduction

Prerequirements:

• Request LLAMA2 access permission and download it , or use other LLAMA2 compatible model, etc. Llama2-Chinese-7b-Chat
• A GPU machine that already install nvidia driver , CUDA , I preferred to use AWS EC2 g5.xlarge instance, or other more than 24G GPU memory instance.
• Some python coding and docker skill .

Chapter 1: Run LLAMA2 model with native python code

This is chapter 1.

Chapter 2

This is chapter 2.

Conclusion

This is the conclusion.