Prompt Engineering In Python, With Gpt, And The Openai Api

Learn to get better answers from LLMs, test the replies, and use the OpenAI Python APi like a pro!

What you'll learn

Turn ChatGPT into a personalised coach

Use the OpenAI Python API to query Large Language Models (LLMs)

Have multi-turn chats with LLMs via the OpenAI Python API

Iteratively improve your prompts

Become aware of prompt hacking and learn how to mitigate it

Extract numeric values from text data

Test your prompts to make sure that they are robust and reliable

Requirements

Option 1: The first chapter of this course uses ChatGPT to teach you Python. So you don't need to know Python if you are comfortable with the idea of learning it from ChatGPT.

Option 2: You may wish to try a dedicated Python course first. However, I would try the ChatGPT option first. Strictly speaking, you should be comfortable using Python in Jupyter notebooks. But you can always enrol in an additional Python course on Udemy.

Description

This course is different to all of the other Prompt Engineering courses. In this course, you will learn then AI skills that you will need if you need if you are working as a software engineer, data scientist, data analyst, or as a real, highly paid, prompt engineer.What is prompt engineering?Prompt Engineering is the skill of finding the right prompt to get the right results from your LLM. With expert level prompt engineering skills, you can implement a commercially profitable LLM solution. On the other hand. A prompt written by someone with average skills, might not lead to a working solution at all. I predict that prompt engineering will become the most demanded skill in the analytics industry within 12 - 24 months. As companies start to take up LLM use cases,  you have an important choice to make. You can either learn these new skills today, and be well placed for the upcoming AI roles. Or you can sit on your hands while other people step into newly created AI roles.How is it different to just using ChatGPT?The bulk of this course teaches you to use the OpenAI Python API to query state of the art LLMs like GPT  3.5 and GPT 4. Do I need to know Python to do this course?No you don't. Because at the start of the course, you will learn to prompt ChatGPT to teach you Python. So if you don't know Python, and you are comfortable with the idea of having "a machine teach you how to program a machine", then go ahead an enroll in this course.What is covered?After learning from our personalised ChatGPT coach, we will start querying GPT with the Python API. You will be able to use either GPT 3.5 Turbo or GPT 4. It's a matter of changing a single parameter.You will learn how to get better answers from LLMs than your untrained competitors. You will also learn the famous "Chain-Of-Thought" prompting technique.Next, you will learn how to have multi-turn conversations with GPT models in Python. The course will also introduce you to prompt hacking. You will see examples of prompt hacking. You will learn how to defend against prompt hacking. And you will see how OpenAI is patching the security holes in its models.Next you will learn how to use the LLM to extract data from text. Then you will learn to properly test the results of your LLM data extraction. This workflow could be a component in a data pipeline. At the end of the course, you will be well on your journey to becoming a professional prompt engineer. What else is unique about this course?There are lots of opportunities to learn on the internet. But when you dig deeper, you will often notice that the presenter lacks commercial experience. The internet is full of training material compiled by recent graduates, undergrad students, sales and marketing teams, and professors who have spent their whole career in academia. The best education however, comes from someone who has real industry experience.You need an experienced professional in your corner. You need someone to show you alternative perspectives that you have never seen before. You need someone to teach you how to “get it”.Who is the instructor and what are his credentials?Hi! My name is Slava Razbash, I have worked in AI and Machine Learning roles since 2011. These roles have been called different names over the years. Names like “Forecasting”, “Predictive Analytics”, “Prescriptive Analytics”, and of course, “Data Science”.Large Language Models are the next step in the evolution of this field. LLMs are already ready offering career opportunities to people with the right skill set. And the number of opportunities will grow as more companies start adopting LLM use cases.You might want to check out my LinkedIn profile. You might want to see where I’ve worked. (You're welcome to connect with me as well.) I’ve solved a lot of data science and machine learning engineering problems. And I have helped a lot of my friends and colleagues. Sometimes ex-colleagues ask me for advice long after I have left the company. But now I’m sharing my experience and knowledge with you - to help you become a highly paid prompt engineer. Now you can learn from me too.Is this course for me?In the realm of career decisions, you hold the reins to your own destiny. The choices that you make today have the potential to shape your future for many years. In terms of learning prompt engineering skills at the professional level, I see four distinct choices.First Choice: Do nothing. Watch the AI boom pass by. If you choose the “do nothing” option, then you will miss out on the career boost from the AI boom. I estimate that the AI boom has the potential to boost your career by five years. Second Choice: Completely switch industries away from anything to do with AI. Maybe you always wanted to become a pilot?Third Choice: Learn Prompt Engineering elsewhere. From instructors who have much less (if any) commercial experience. It will take longer. The courses and books will not be tailored to commercial use cases. You will spend lots of time writing prompts about counting apples and other trivial applications. You will have to figure out how to transfer your new generic knowledge to the corporate setting. You also won’t learn my secret techniques. Fourth Choice: Take this course, and give yourself a career boost. Because this is the fastest path to LLM mastery in the corporate data analytics space.You’re not confined to a single path. The choice is in your hands!

Overview

Section 1: Introduction

Lecture 1 Course Overview and Key Points of Difference

Lecture 2 Introduction

Section 2: ChatGPT as a Personal Coach

Lecture 3 Turning ChatGPT into a Personal Coach

Lecture 4 Where ChatGPT fails

Section 3: Notebook 1: Python API Prompting

Lecture 5 Setting up your OpenAI API Key

Lecture 6 Python API Prompting and Iterative Prompt Improvement Part 1

Lecture 7 Iterative Prompting Improvement Part 2

Lecture 8 Using GPT 4 and Other Model Versions

Lecture 9 Optional: Implementing Exponential Backoff for Querying the OpenAI API

Lecture 10 Inducing and Mitigating Hallucination in LLMs

Lecture 11 Chain-Of-Thought Prompting - the formal introduction

Section 4: Python API Chat Bot

Lecture 12 Python API Chat Bot

Lecture 13 Prompt Hacking

Lecture 14 Optional: Implementing Exponential Backoff for Multi-Turn Conversations

Section 5: Extracting Numerical Values from Text Data

Lecture 15 Extracting Numerical Values from Text Data, and Testing the Extraction.

People who wish to pursue a career in as a professional Prompt Engineer,People analyse data in a corporate role,People who wish to pursue a career in data analytics, data science, or any other field that analyses data