ChatGPT as a Research Assistant? – Annotate your data with Language Models

Thursday, March 21 – 10:00-12:00

Room 1.1.12

To RSVP, email datalab@samf.ku.dk. Please also email us with suggestions on skills and programs you’d like to see covered in a workshop.

A hands-on workshop on annotating your data using off-the-self LLMs. We will mainly focus on text and image data.

We will start with interacting with the OpenAI API (setting up an account, pricing, models available, prompt engineering, overcoming issues with large prompts, best practices, approximating deterministic-replicable outputs etc). We will then focus on setting up an open-source model from hugging face locally as an alternative to OpenAI’s API and we will compare the results.

If participants are interested, we might also dedicate some time to setting up an account and script to access Deepl’s API for translation.

Some experience with Python and working with APIs is needed. This is not geared towards beginners but if you are not sure if you can make the most out of the workshop, feel free to reach out to us to discuss this. If we get enough interest from researchers, we may hold a ChatGPT as a RA for beginners' course as well. 

Additionally, KUB Data Lab offers a series of Python for absolute beginners workshops. We discussed this with KUB Data Lab and you are more than welcome to enroll ahead of our workshop if there are empty spots left.

To RSVP, email datalab@samf.ku.dk. Please also email us with suggestions on skills and programs you’d like to see covered in a workshop.