Tutorial: ChatGPT, Generative AI and Prompt Engineering for Industrial Applications

Welcome to the tutorial on ChatGPT, Generative AI and Prompt Engineering for Industrial Applications at the IEEE-ISIE 2023 conference. This tutorial is organized by Daswin De Silva, Nishan Mills, and Gihan Gamage from La Trobe University, Victoria, Australia.

The tutorial consists of three Jupyter notebooks that provide a comprehensive guide to the theory and practice of using generative AI models and libraries for the development of industrial applications and solutions.

Notebooks Included

1. Transcription and Query Interface for YouTube Videos: This notebook takes a YouTube video URL, transcribes it using OpenAI's Whisper ASR API, and then vectorizes and stores the transcription in a ChromaDB vector database. It then uses Langchain to provide a query interface to the vector database.

2. CSV Data Analysis and Graph Generation: This notebook reads a CSV file and generates Python code to create a graph based on prompts.

3. Google Drive File Analysis and Vector Storage: This notebook looks at a folder in Google Drive, chunks and stores the embeddings made with OpenAI's DAVinci API in a ChromaDB vector database, and then queries the database using Langchain.

4. Prompt Engineering Basics: This notebook focuses on basics in prompt engineering such as parameters in llms and prompt engineering techniques.

All notebooks are designed to be used with Google Colab, a cloud-based Jupyter notebook environment that requires no setup and runs entirely in the cloud.

Prerequisites

1. An API Key from OpenAI; you can create an OpenAI API key for free. New free trial users receive $5 (USD) worth of credit, which expires after three months. Once your credit has been used up or expired, you can enter billing information to continue using the API. Accounts are unique across phone numbers.

2. A google account with which to access Google Colab.

Setup Instructions

Option 1: Open Directly in Google Colab

1. Click on the Google Colab link provided in the repository to open the notebook.

Option 2: Load from Github into Google Colab

1. Go to Google Colab.
2. Click on the File menu, then select Open notebook.
3. In the dialog box that opens, select the GitHub tab.
4. In the search box, paste the GitHub repository URL and press Enter.
5. From the list of notebooks that appear, click on the one you want to open.

Option 3: Download from Github and Open in Google Colab

1. Navigate to the main page of the repository.
2. Click on the Code button which is usually towards the right of the page.
3. Click on Download ZIP.
4. Extract the ZIP file in your local system.
5. Go to Google Colab.
6. Click on File -> Open notebook.
7. In the dialog box that opens, select the Upload tab.
8. Click on Choose File, navigate to the extracted folder location, and open the .ipynb file.

Note

• You need to have a Google account to use Google Colab.
• If you are using Google Colab for the first time, you may need to authorize it with your Google account.

About Google Colaboratory

Google Colaboratory (Colab) is a free cloud-based Jupyter notebook environment that allows you to write and execute Python code with zero configuration required. Colab comes with robust integration with Google Drive, GitHub, and many popular machine learning libraries, making your data analysis, machine learning experimentation, or simple Python scripting as seamless as possible. All computations take place on Google's hardware, allowing you to leverage the power of Google's infrastructure, including GPUs and TPUs. You can also share, comment, and collaborate on any of your Colab notebooks, just as you would with Google Docs.

You can find out more about how you can make use of Google Colab here.

About the Tutorial

The tutorial will cover the theory and practice of using generative AI models and libraries for the development of industrial applications and solutions. It begins by exploring the structural elements of Generative AI models, transformers, hyper-parameters, transfer learning and comparison to standard machine learning algorithms. It then moves on to the application of generative AI for the design, development and evaluation of industrial applications. Participants will develop hands-on skills in using generative AI libraries and acquire a practical understanding of “prompt engineering” for diverse industrial settings. The learning outcomes of this workshop are:

• Theoretical foundations of Generative AI - when to use and in which settings
• Design and development of Generative AI models
• Prompt engineering for diverse use cases
• Rapid prototyping to evaluation of a suitable Generative AI solution

Enjoy the tutorial and happy learning!