http://arxiv.org/api/GrfVMsjRJ5QfXT4borMr9RuzWhk 2023-06-09T00:00:00-04:00 1 0 1 http://arxiv.org/abs/1910.01108v4 2020-03-01T02:57:50Z 2019-10-02T17:56:28Z

As Transfer Learning from large-scale pre-trained models becomes more prevalent in Natural Language Processing (NLP), operating these large models in on-the-edge and/or under constrained computational training or inference budgets remains challenging. In this work, we propose a method to pre-train a smaller general-purpose language representation model, called DistilBERT, which can then be fine-tuned with good performances on a wide range of tasks like its larger counterparts. While most prior work investigated the use of distillation for building task-specific models, we leverage knowledge distillation during the pre-training phase and show that it is possible to reduce the size of a BERT model by 40%, while retaining 97% of its language understanding capabilities and being 60% faster. To leverage the inductive biases learned by larger models during pre-training, we introduce a triple loss combining language modeling, distillation and cosine-distance losses. Our smaller, faster and lighter model is cheaper to pre-train and we demonstrate its capabilities for on-device computations in a proof-of-concept experiment and a comparative on-device study.

Victor Sanh Lysandre Debut Julien Chaumond Thomas Wolf February 2020 - Revision: fix bug in evaluation metrics, updated metrics, argumentation unchanged. 5 pages, 1 figure, 4 tables. Accepted at the 5th Workshop on Energy Efficient Machine Learning and Cognitive Computing - NeurIPS 2019