BERT for Humanists Virtual Workshop
Friday June 4th, 2021 — 1pm-3pm Eastern
Hosted by David Mimno, Melanie Walsh, and Maria Antoniak
You can register for the Zoom workshop here
In the last three years, natural language processing researchers have been abuzz about a state-of-the-art method called BERT…
What is BERT? How do you use it? What kinds of digital humanities projects will BERT be most useful for?
This free, two-hour virtual workshop will begin to answer these important questions about BERT-like models and their potential role in DH research and teaching.
We will first offer a brief conceptual overview of BERT. Then we will dive into a hands-on, code-based tutorial that demonstrates how to train and fine-tune a BERT model with HuggingFace’s popular Python library
transformers. This demonstration will be focused on humanistic data, including poetry and Goodreads book reviews.
Who is this workshop for?
This workshop is primarily geared toward digital humanists and computational social scientists who are familiar with Python and have a basic knowledge of machine learning. However, all participants with an interest in the humanities and social sciences are welcome.
We would like to emphasize that this is not a general-purpose BERT workshop.
Where do I sign up?
You can register for the Zoom workshop here.
What can I do to prepare?
You might review some of the basic machine learning concepts and tools that we will drawn upon during the workshop, such as training and testing and the Python library
- “The 7 Steps of Machine Learning,” Yufeng Guo
[YouTube video][Blog post]
- “A Beginners Guide to Scikit-Learn”, Rebecca Vickery
You might also familiarize yourself with Google Colab notebooks, which will be using extensively throughout the workshop:
What if I can’t make it?
We’re sorry to miss you! We plan to release selected recordings from the workshop on YouTube.