paper-with-me

Papers

Sparse Progressive Distillation: Resolving Overfitting under Pretrain-and-Finetune Paradigm

2021-10-15 · ACL 2022 5 · Shaoyi Huang, Dongkuan Xu, Ian E. H. Yen, Yijue Wang, Sung-En Chang, Bingbing Li, Shiyang Chen, Mimi Xie, Sanguthevar Rajasekaran, Hang Liu, Caiwen Ding

Conventional wisdom in pruning Transformer-based language models is that pruning reduces the model expressiveness and thus is more likely to underfit rather than overfit. However, under the trending pretrain-and-finetune paradigm, we postulate a counter-traditional hypothesis, that is: pruning increases the risk of overfitting when performed at the fine-tuning phase. In this paper, we aim to address the overfitting problem and improve pruning performance via progressive knowledge distillation with error-bound properties. We show for the first time that reducing the risk of overfitting can help the effectiveness of pruning under the pretrain-and-finetune paradigm. Ablation studies and experiments on the GLUE benchmark show that our method outperforms the leading competitors across different tasks.

📄 PDF Abstract BibTeX arXiv:2110.08190

Code (0)

등록된 구현이 없습니다.

Tasks

Knowledge Distillation

Methods 이 논문이 사용한 방법론

Pruning 설명 없음
Knowledge Distillation A very simple way to improve the performance of almost any machine learning algorithm is to train many different models on the same data and then to average their predictions.…

Similar Papers 제목 키워드 기반

Sparse Progressive Distillation: Resolving Overfitting under Pretrain-and-Finetune Paradigm

2021-11-16 · ACL ARR November 2021 11 · Anonymous

Conventional wisdom in pruning Transformer-based language models is that pruning reduces the model expressiveness and thus is more likely to underfit rather than overfit. However, under the trending pretrain-and-finetune…

Knowledge Distillation

Efficient Knowledge Distillation via Curriculum Extraction

2025-03-21 · Shivam Gupta, Sushrut Karmalkar

Knowledge distillation is a technique used to train a small student network using the output generated by a large teacher network, and has many empirical advantages~\citep{Hinton2015DistillingTK}. While the standard one-…

Knowledge DistillationLanguage ModelingLanguage Modelling

Progressive distillation induces an implicit curriculum

2024-10-07 · Abhishek Panigrahi, Bingbin Liu, Sadhika Malladi, Andrej Risteski 외

Knowledge distillation leverages a teacher model to improve the training of a student model. A persistent challenge is that a better teacher does not always yield a better student, to which a common mitigation is to use …

Knowledge Distillation

Advancing Brain Imaging Analysis Step-by-step via Progressive Self-paced Learning

2024-07-23 · Yanwu Yang, Hairui Chen, Jiesi Hu, Xutao Guo 외

Recent advancements in deep learning have shifted the development of brain imaging analysis. However, several challenges remain, such as heterogeneity, individual variations, and the contradiction between the high dimens…

Improving Generalization of Metric Learning via Listwise Self-distillation

2022-06-17 · Zelong Zeng, Fan Yang, Zheng Wang, Shin'ichi Satoh

Most deep metric learning (DML) methods employ a strategy that forces all positive samples to be close in the embedding space while keeping them away from negative ones. However, such a strategy ignores the internal rela…

Metric Learning