paper-with-me

Papers

Teacher Intervention: Improving Convergence of Quantization Aware Training for Ultra-Low Precision Transformers

2023-02-23 · Minsoo Kim, Kyuhong Shim, Seongmin Park, Wonyong Sung, Jungwook Choi

Pre-trained Transformer models such as BERT have shown great success in a wide range of applications, but at the cost of substantial increases in model complexity. Quantization-aware training (QAT) is a promising method to lower the implementation cost and energy consumption. However, aggressive quantization below 2-bit causes considerable accuracy degradation due to unstable convergence, especially when the downstream dataset is not abundant. This work proposes a proactive knowledge distillation method called Teacher Intervention (TI) for fast converging QAT of ultra-low precision pre-trained Transformers. TI intervenes layer-wise signal propagation with the intact signal from the teacher to remove the interference of propagated quantization errors, smoothing loss surface of QAT and expediting the convergence. Furthermore, we propose a gradual intervention mechanism to stabilize the recovery of subsections of Transformer layers from quantization. The proposed schemes enable fast convergence of QAT and improve the model accuracy regardless of the diverse characteristics of downstream fine-tuning tasks. We demonstrate that TI consistently achieves superior accuracy with significantly lower fine-tuning iterations on well-known Transformers of natural language processing as well as computer vision compared to the state-of-the-art QAT methods.

📄 PDF Abstract BibTeX arXiv:2302.11812

Code (1)

marsjacobs/ti-kd-qat 공식 구현 pytorch

Tasks

Knowledge DistillationQuantization

Methods 이 논문이 사용한 방법론

Attention 설명 없음
Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
Weight Decay 설명 없음
Adam 설명 없음
Multi-Head Attention 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.
Attention Dropout Attention Dropout is a type of dropout used in attention-based architectures, where elements are randomly dropped out of the…
Residual Connection 설명 없음

Similar Papers 제목 키워드 기반

PQK: Model Compression via Pruning, Quantization, and Knowledge Distillation

2021-06-25 · Jangho Kim, Simyung Chang, Nojun Kwak

As edge devices become prevalent, deploying Deep Neural Networks (DNN) on edge devices has become a critical issue. However, DNN requires a high computational resource which is rarely available for edge devices. To handl…

Keyword SpottingKnowledge DistillationModel CompressionQuantization

QKD: Quantization-aware Knowledge Distillation

2019-11-28 · Jangho Kim, Yash Bhalgat, Jinwon Lee, Chirag Patel 외

Quantization and Knowledge distillation (KD) methods are widely used to reduce memory and power consumption of deep neural networks (DNNs), especially for resource-constrained edge devices. Although their combination is …

Knowledge DistillationQuantization

Efficient INT8 Single-Image Super-Resolution via Deployment-Aware Quantization and Teacher-Guided Training

2026-04-22 · Pham Phuong Nam Nguyen, Nam Tien Le, Thi Kim Trang Vo, Nhu Tinh Anh Nguyen arxiv

Efficient single-image super-resolution (SISR) requires balancing reconstruction fidelity, model compactness, and robustness under low-bit deployment, which is especially challenging for x3 SR. We present a deployment-or…

Image Super-Resolution

Q$^2$: Quantization-Aware Gradient Balancing and Attention Alignment for Low-Bit Quantization

2025-11-08 · Zhaoyang Wang, Dong Wang arxiv

Quantization-aware training (QAT) has achieved remarkable success in low-bit ($\leq$4-bit) quantization for classification networks. However, when applied to more complex visual tasks such as object detection and image s…

Image SegmentationObject Detection

AWP: Activation-Aware Weight Pruning and Quantization with Projected Gradient Descent

2025-06-11 · Jing Liu, Toshiaki Koike-Akino, Ye Wang, Hassan Mansour 외

To address the enormous size of Large Language Models (LLMs), model compression methods, such as quantization and pruning, are often deployed, especially on edge devices. In this work, we focus on layer-wise post-trainin…

Model CompressionQuantization