T3-Vis: a visual analytic framework for Training and fine-Tuning Transformers in NLP
This is an incremental tool for NLP researchers to better understand and optimize Transformer models.
The paper tackles the challenge of training and fine-tuning Transformers in NLP by introducing a visual analytic framework that provides insights into model properties and behaviors, with case studies and user feedback indicating its usefulness.
Transformers are the dominant architecture in NLP, but their training and fine-tuning is still very challenging. In this paper, we present the design and implementation of a visual analytic framework for assisting researchers in such process, by providing them with valuable insights about the model's intrinsic properties and behaviours. Our framework offers an intuitive overview that allows the user to explore different facets of the model (e.g., hidden states, attention) through interactive visualization, and allows a suite of built-in algorithms that compute the importance of model components and different parts of the input sequence. Case studies and feedback from a user focus group indicate that the framework is useful, and suggest several improvements.