M&S Research Hay Day 2023 Friday 2nd of June 2023 Repository
Welcome to Introduction to Transformers, BERT, and Explainable NLP
Instructors:
Huyen Nguyen, Daniel Anadria, Hadi Mohammadi
Description:
In this workshop, we will delve into the world of transformer-based language models, focusing on the BERT model, which has revolutionized natural language processing tasks (Vaswani et al., 2017; Devlin et al., 2018).
We will illustrate the application of these models in sentiment analysis, demonstrating how the BERT model can assess the emotional tone of the text. Further, we will introduce SHAP, a game-theoretic approach that elucidates the influence of specific words or phrases on the predictions made by these models (Lundberg & Lee, 2017). All the required materials, including the code, intuitive explanations and further reading materials, are shared via this GitHub repository.
Preparation: This interactive session will be conducted on Google Colab, requiring nothing more than a laptop, a Gmail account with an internet connection for participation.