paper-with-me

홈 › Papers

Improving Task-Agnostic BERT Distillation with Layer Mapping Search

2020-12-11 · Xiaoqi Jiao, Huating Chang, Yichun Yin, Lifeng Shang, Xin Jiang, Xiao Chen, Linlin Li, Fang Wang, Qun Liu

Knowledge distillation (KD) which transfers the knowledge from a large teacher model to a small student model, has been widely used to compress the BERT model recently. Besides the supervision in the output in the original KD, recent works show that layer-level supervision is crucial to the performance of the student BERT model. However, previous works designed the layer mapping strategy heuristically (e.g., uniform or last-layer), which can lead to inferior performance. In this paper, we propose to use the genetic algorithm (GA) to search for the optimal layer mapping automatically. To accelerate the search process, we further propose a proxy setting where a small portion of the training corpus are sampled for distillation, and three representative tasks are chosen for evaluation. After obtaining the optimal layer mapping, we perform the task-agnostic BERT distillation with it on the whole corpus to build a compact student model, which can be directly fine-tuned on downstream tasks. Comprehensive experiments on the evaluation benchmarks demonstrate that 1) layer mapping strategy has a significant effect on task-agnostic BERT distillation and different layer mappings can result in quite different performances; 2) the optimal layer mapping strategy from the proposed search process consistently outperforms the other heuristic ones; 3) with the optimal layer mapping, our student model achieves state-of-the-art performance on the GLUE tasks.

📄 PDF Abstract BibTeX arXiv:2012.06153

Code (0)

등록된 구현이 없습니다.

Tasks

Knowledge Distillation

Methods 이 논문이 사용한 방법론

Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
Residual Connection 설명 없음
Linear Warmup With Linear Decay Linear Warmup With Linear Decay is a learning rate schedule in which we increase the learning rate linearly for $n$ updates and then linearly decay afterwards.
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 &…
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…
Attention 설명 없음
Layer Normalization Unlike batch normalization, Layer Normalization directly estimates the normalization statistics from the summed inputs…
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…

Similar Papers 제목 키워드 기반

Improving Knowledge Distillation for BERT Models: Loss Functions, Mapping Methods, and Weight Tuning

2023-08-26 · Apoorv Dankar, Adeem Jassani, Kartikaeya Kumar

The use of large transformer-based models such as BERT, GPT, and T5 has led to significant advancements in natural language processing. However, these models are computationally expensive, necessitating model compression…

Knowledge DistillationModel Compression

How to Distill your BERT: An Empirical Study on the Impact of Weight Initialisation and Distillation Objectives

2023-05-24 · Xinpeng Wang, Leonie Weissweiler, Hinrich Schütze, Barbara Plank

Recently, various intermediate layer distillation (ILD) objectives have been shown to improve compression of BERT models via Knowledge Distillation (KD). However, a comprehensive evaluation of the objectives in both task…

Knowledge DistillationQNLI

BERT-EMD: Many-to-Many Layer Mapping for BERT Compression with Earth Mover's Distance

2020-10-13 · EMNLP 2020 11 · Jianquan Li, Xiaokang Liu, Honghong Zhao, Ruifeng Xu 외

Pre-trained language models (e.g., BERT) have achieved significant success in various natural language processing (NLP) tasks. However, high storage and computational costs obstruct pre-trained language models to be effe…

Model Compression

Universal-KD: Attention-based Output-Grounded Intermediate Layer Knowledge Distillation

2021-11-01 · EMNLP 2021 11 · Yimeng Wu, Mehdi Rezagholizadeh, Abbas Ghaddar, Md Akmal Haidar 외

Intermediate layer matching is shown as an effective approach for improving knowledge distillation (KD). However, this technique applies matching in the hidden spaces of two different networks (i.e. student and teacher),…

Knowledge Distillation

Weight-Inherited Distillation for Task-Agnostic BERT Compression

2023-05-16 · Taiqiang Wu, Cheng Hou, Shanshan Lao, Jiayi Li 외

Knowledge Distillation (KD) is a predominant approach for BERT compression. Previous KD-based methods focus on designing extra alignment losses for the student model to mimic the behavior of the teacher model. These meth…

Knowledge Distillation