학술논문
MetaBERT: Collaborative Meta-Learning for Accelerating BERT Inference
Document Type
Conference
Author
Source
2023 26th International Conference on Computer Supported Cooperative Work in Design (CSCWD) Computer Supported Cooperative Work in Design (CSCWD), 2023 26th International Conference on. :119-124 May, 2023
Subject
Language
ISSN
2768-1904
Abstract
Early exit methods are used to accelerate inference in pre-trained language models and maintain competitive performance on resource-constrained devices. However, existing methods for training early exit classifiers suffer from the problem of poor classifier representations in different layers, leading to difficulties in adapting to diverse natural language processing tasks. To address this issue, we propose MetaBERT: collaborative Meta-learning for accelerating BERT inference. The main goal of MetaBERT is to train early exit classifiers through collaborative meta-learning, in which case, few gradient updates can be quickly adapted to new tasks. Moreover, this novel meta-training approach produces good generalization performance, thus achieving an effective balance between the inference result and efficiency. Extensive experimental results show that our approach outperforms previous training methods by a large margin, and achieves state-of-the-art results compared to other competitive models.