OpenAI Workshop

This is accompanying information for a mini workshop at tretton37.

The Goal of this Workshop

The primary aim of this workshop is to spark creativity by demonstrating the potential applications of the OpenAI API.

We will drive this learning journey by focusing on the development of a Git commit suggestion tool, showcasing practical implementation of the OpenAI API.

Prerequisites

• Git
• OpenAI account

Workshop Structure

• Introduction
• Coding Session 1
• OpenAI Details
• Coding Session 2
• Wrap Up

Introduction

An overview of OpenAI and its various AI models, emphasizes caution with proprietary data, explains the pay-as-you-go pricing model, mentions available programming libraries, and plans to detail strategies for interacting with the tool.

👉 Introduction 👈

Coding Session 1

Let's get the ball rolling by getting an OpenAI key, devise a strategy for integrating git diffs into their tool, and enable the tool to send requests to the OpenAI API

👉 Coding Session 1 👈

OpenAI API Overview

A quick insight on how to interact with the OpenAI API. Covers 'Authorization', the chosen model, and a series of messages in the JSON payload. We explore important parameters like 'temperature' that can influence the output's randomness, and touch on the concept of tokens - central to usage and cost implications.

Lastly, we delve into unique characteristics of the models including their knowledge cutoff, potential for fabricating information, and their imperfect word counting abilities. Strategies for effective git diff and commit as well as tips for formulating compelling prompts also form part of the tutorial, enhancing your overall proficiency with OpenAI's GPT-4 API.

👉 OpenAI Overview 👈

Coding Session 2

Lets get going with creating prompts and using this in our tool! After this you should have a finished CLI tool to create commit message suggestions!

👉 Coding Session 2 👈

Wrap Up

Proposed improvements, questions and further resources.

👉 Wrap Up 👈

Further Resources

• openai.com Documentation

• ChatGPT Prompt Engineering for Developers

• OpenAI Playground https://platform.openai.com/playground/p/?mode=chat

• OpenAI Cookbook: https://github.com/openai/openai-cookbook/blob/main/techniques_to_improve_reliability.md

Examples are available in the examples folder.