Large Language Models for Digital Humanities: A Hands-On Introduction

Large Language Models (LLMs) are increasingly used in both research and applied settings, including Digital Humanities (DH) in the broadest sense. To work with them effectively, researchers must understand not only their conceptual foundations but also the practical aspects of deployment, inference, and evaluation.

This hands-on workshop introduces participants to the practical use of LLMs in research workflows. They will learn how to access and run modern LLMs locally or on GPU servers, configure them for efficient inference, and apply them to common text-based tasks. The workshop assumes basic familiarity with Python but no prior experience with machine-learning frameworks.

The workshop combines theoretical overviews with an emphasis on guided practical exercises. By the end of the workshop, participants will be able to set up LLM-based experiments and critically assess their suitability for DH research.

Topics covered include:

1. Essential background concepts for working with LLMs, including machine learning basics, text representation, transformer models, and inference strategies;
2. Installing and running LLMs locally, as well as accessing models via APIs on local or remote GPU servers;
3. Using LLMs for text classification and related analytical tasks;
4. Advanced prompting techniques, including Retrieval-Augmented Generation (RAG), to improve text generation;
5. Practical use of state-of-the-art inference libraries for efficient and scalable experimentation;
6. Individual mentoring and support during hands-on exercises and exploratory tasks.