TextBrewer: An Open-Source Knowledge Distillation Toolkit for Natural Language Processing
This provides an open-source tool for NLP researchers and practitioners to easily implement knowledge distillation, though it is incremental as it builds on existing distillation methods.
The authors tackled the need for a flexible knowledge distillation toolkit in NLP by introducing TextBrewer, which achieved results comparable to or higher than public distilled BERT models on typical tasks with similar parameter counts.
In this paper, we introduce TextBrewer, an open-source knowledge distillation toolkit designed for natural language processing. It works with different neural network models and supports various kinds of supervised learning tasks, such as text classification, reading comprehension, sequence labeling. TextBrewer provides a simple and uniform workflow that enables quick setting up of distillation experiments with highly flexible configurations. It offers a set of predefined distillation methods and can be extended with custom code. As a case study, we use TextBrewer to distill BERT on several typical NLP tasks. With simple configurations, we achieve results that are comparable with or even higher than the public distilled BERT models with similar numbers of parameters. Our toolkit is available through: http://textbrewer.hfl-rc.com