We will be exploring how to use AI in our data analysis pipeline.
For this workshop, I have set up several examples to work with, canvassing different possibilities. You are welcome to change the code to suit your purposes.
Minimal understanding of programming required. This repository is intended to be an accessible starting point to non-experts.
Google Drive
Google Colaboratory -- To check whether you have this, see if you can choose it as an option for creating a new document in the online Google Drive interface (New > More), otherwise click "Connect More Apps".
If you wish to run the entire script for the second file, you need an OpenAI platform account; should come with $5 sign on bonus. It will not run if you don't have any money on the platform.
Download the ZIP file of this repository (top right, green button that says Code), and drop the files into your Google Drive, in a place where you can easily find it.
Please start by opening scripts folder, and click on the text-insights...ipynb script file. At this point, you should be in the Google Colaboratory environment.
By pressing the "play" button next to each bit of code, you can run the Python code, and it will usually present you with some output.
After you get comfortable, try out the translate-and-code...ipynb script.
If you change the code and use some of your own data, note that Google Colab is connected to Google servers and that if you use the OpenAI API, you will be sending your data to them.
The HuggingFace models work differently - you load the model into your RAM and pass it requests.
Should you have any problems, please send an email to either my permanent email: simonvanbaal.behsci@gmail.com, or my institutional email listed on my profile page.
This project is not open for contributions. However, if you find a mistake, please contact me via the email address listed above.