l1-intro-notes.md

Lesson 1: Introduction

Overview

• A lot of material on the internet for prompting has been focused on the chatGPT web user interface, which many people are using to do specific and often one-off tasks
• The power of LLMs (Large Language Models), as a developer tool, that is using API calls
• LLM APIs can enable developers to very quickly build software applications

Course Plan

• Prompting best practices for software development
• Some common use cases :
  • summarizing
  • inferring
  • transforming
  • expanding
• Build a chatbot using an LLM

Two types of LLMs

• Base LLMs : predicts next words, based on text training data (huge amount of text data from internet and another sources) to figure it out most likely word to follow
  • Ex : what is the capital of France?
    • What is France's largest city?
    • What is France's population?
    • What is the currency of France?
• Instruction Tuned LLMs :

  • Tries to follow instructions.

  • Fine-tune on instructions(input/outputs) and good attempts at following those instructions

  • Uses Reinforcement Learning with Human Feedback (RLHF) technique

  • Trained to be : Helpful, Honest , Harmless, safe*

    • Ex: what is the capital of France?
      • The capital of France is Paris

TL;DR

The 6 Prompt engineering strategies to get the best results from Large Language Models (LLMs):

1. Write clear instructions
2. Provide reference text
3. Split complex tasks into simpler subtasks
4. Give the model time to "think”
5. Use external tools
6. Test changes systematically

References

Main Course :

• https://learn.deeplearning.ai/chatgpt-prompt-eng/lesson/1/introduction

Model index for researchers - openai:

• https://platform.openai.com/docs/model-index-for-researchers/model-index-for-researchers

Generative Pre-trained Transformer Models :

• GPT-1 : https://en.wikipedia.org/wiki/Generative_pre-trained_transformer

• GPT-2 : https://en.wikipedia.org/wiki/GPT-2

  • paper : https://arxiv.org/ftp/arxiv/papers/1908/1908.09203.pdf

• GPT-3 : https://en.wikipedia.org/wiki/GPT-3

  • paper : https://arxiv.org/pdf/2005.14165.pdf

• GPT-4 : https://en.wikipedia.org/wiki/GPT-4

  • paper : https://arxiv.org/pdf/2303.08774.pdf

• Codex :

  • paper : https://arxiv.org/pdf/2107.03374.pdf

• InstructGPT (ChatGPT ancestor) :

  • paper : https://arxiv.org/pdf/2203.02155.pdf

GPT Best Practices - openai :

• https://platform.openai.com/docs/guides/gpt-best-practices

What are Generative AI models? - IBM

• https://www.youtube.com/watch?v=hfIUstzHs9A

Cohere - Transform your products with LLMs : https://cohere.com/

Langchains : https://python.langchain.com/docs/get_started/introduction.html

huggingface - The AI community building the future - https://huggingface.co/

anthropic - AI research and products that put safety at the frontier https://www.anthropic.com/