paper-with-me

Papers

Feature Structure Distillation for BERT Transferring

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

Knowledge distillation is an approach to transfer information on feature representations from a teacher to a student by reducing their difference. A challenge of this approach is to reduce the flexibility of the student's representations inducing inaccurate learning of the teacher's knowledge. To resolve it in BERT transferring, we investigate distillation of structures of representations specified to three types: intra-feature, local inter-feature, global inter-feature structures. To transfer them, we introduce \textit{feature structure distillation} methods based on the Centered Kernel Alignment, which assigns a consistent value to similar distributions of representations and reveals more informative relations. In particular, a memory-augmented transfer method with clustering is implemented for the global structures. In the experiments on the nine tasks for language understanding of the GLUE dataset, the proposed methods effectively transfer the three types of structures and improve performance compared to state-of-the-art distillation methods.

📄 PDF Abstract BibTeX

Code (0)

등록된 구현이 없습니다.

Tasks

Knowledge Distillation

Methods 이 논문이 사용한 방법론

Refunds@Expedia|||How do I get a full refund from Expedia? “How do I get a full refund from Expedia? How do I get a full refund from Expedia? – Call ☎️ +1-(888) 829 (0881) or +1-805-330-4056 or +1-805-330-4056 for Quick Help &…
Attention 설명 없음
Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
Dropout Dropout is a regularization technique for neural networks that drops a unit (along with connections) at training time with a specified probability $p$ (a common value is…
Weight Decay 설명 없음
Multi-Head Attention 설명 없음
Dense Connections Dense Connections, or Fully Connected Connections, are a type of layer in a deep neural network that use a linear operation where every input is connected to every output…
Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…

Similar Papers 제목 키워드 기반

Feature Structure Distillation with Centered Kernel Alignment in BERT Transferring

2022-04-01 · Hee-Jun Jung, Doyeon Kim, Seung-Hoon Na, Kangil Kim

Knowledge distillation is an approach to transfer information on representations from a teacher to a student by reducing their difference. A challenge of this approach is to reduce the flexibility of the student's repres…

Knowledge DistillationLanguage ModelingLanguage Modelling

Knowledge distillation through geometry-aware representational alignment

2025-09-27 · Prajjwal Bhattarai, Mohammad Amjad, Dmytro Zhylko, Tuka Alhanai arxiv

Knowledge distillation is a common paradigm for transferring capabilities from larger models to smaller ones. While traditional distillation methods leverage a probabilistic divergence over the output of the teacher and …

Knowledge Distillation

LightMBERT: A Simple Yet Effective Method for Multilingual BERT Distillation

2021-03-11 · Xiaoqi Jiao, Yichun Yin, Lifeng Shang, Xin Jiang 외

The multilingual pre-trained language models (e.g, mBERT, XLM and XLM-R) have shown impressive performance on cross-lingual natural language understanding tasks. However, these models are computationally intensive and di…

Natural Language UnderstandingXLM-R

Towards Non-task-specific Distillation of BERT via Sentence Representation Approximation

2020-04-07 · Asian Chapter of the Association for Computational Linguistics 2020 · Bowen Wu, huan zhang, Mengyuan Li, Zongsheng Wang 외

Recently, BERT has become an essential ingredient of various NLP deep models due to its effectiveness and universal-usability. However, the online deployment of BERT is often blocked by its large-scale parameters and hig…

Knowledge DistillationSentenceTransfer Learning

Neural Collapse Inspired Knowledge Distillation

2024-12-16 · Shuoxi Zhang, Zijian Song, Kun He

Existing knowledge distillation (KD) methods have demonstrated their ability in achieving student network performance on par with their teachers. However, the knowledge gap between the teacher and student remains signifi…

Knowledge Distillation