Accelerating Pretrained Language Model Inference Using Weighted Ensemble Self-distillation
Authors:
- 6 October 2021
- book chapter
- Published by Springer Nature
- p. 224-235
Abstract
No abstract availableKeywords
This publication has 9 references indexed in Scilit:
- Learning sentiment sentence representation with multiview attention modelInformation Sciences, 2021
- Personalized sentiment classification of customer reviews via an interactive attributes attention modelKnowledge-Based Systems, 2021
- The Right Tool for the Job: Matching Model and Instance Complexities2020
- PLATO: Pre-trained Dialogue Generation Model with Discrete Latent Variable2020
- DeeBERT: Dynamic Early Exiting for Accelerating BERT Inference2020
- Q8BERT: Quantized 8Bit BERT2019
- Patient Knowledge Distillation for BERT Model Compression2019
- What Does BERT Learn about the Structure of Language?2019
- BranchyNet: Fast inference via early exiting from deep neural networks2016