Greedy-layer pruning: Speeding up transformer models for natural language processing
Authors:
- David PeerDPDavid Peer
- Sebastian StabingerSSSebastian Stabinger
- Stefan EnglSEStefan Engl
- Antonio Rodríguez-SánchezARAntonio Rodríguez-Sánchez
- 27 March 2022
- journal article
- research article
- Published by Elsevier in Pattern Recognition Letters
- Vol. 157, 76-82
Abstract
No abstract availableKeywords
All Related Versions
This publication has 8 references indexed in Scilit:
- Compressing Large-Scale Transformer-Based Models: A Case Study on BERTTransactions of the Association for Computational Linguistics, 2021
- Document-level relation extraction via graph transformer networks and temporal convolutional networksPattern Recognition Letters, 2021
- Limitation of capsule networksPattern Recognition Letters, 2021
- How to Train BERT with an Academic Budget2021
- Know what you don't need: Single-Shot Meta-Pruning for attention headsAI Open, 2021
- Multi-task learning for natural language processing in the 2020s: Where are we going?Pattern Recognition Letters, 2020
- CapsNet topology to classify tumours from brain images and comparative evaluationIET Image Processing, 2020
- Diagnosis of Alzheimer's disease with Sobolev gradient‐based optimization and 3D convolutional neural networkInternational Journal for Numerical Methods in Biomedical Engineering, 2019