paper-with-me

홈 › Papers

Block-Wise Dynamic-Precision Neural Network Training Acceleration via Online Quantization Sensitivity Analytics

2022-10-31 · Ruoyang Liu, Chenhan Wei, Yixiong Yang, Wenxun Wang, Huazhong Yang, Yongpan Liu

Data quantization is an effective method to accelerate neural network training and reduce power consumption. However, it is challenging to perform low-bit quantized training: the conventional equal-precision quantization will lead to either high accuracy loss or limited bit-width reduction, while existing mixed-precision methods offer high compression potential but failed to perform accurate and efficient bit-width assignment. In this work, we propose DYNASTY, a block-wise dynamic-precision neural network training framework. DYNASTY provides accurate data sensitivity information through fast online analytics, and maintains stable training convergence with an adaptive bit-width map generator. Network training experiments on CIFAR-100 and ImageNet dataset are carried out, and compared to 8-bit quantization baseline, DYNASTY brings up to $5.1\times$ speedup and $4.7\times$ energy consumption reduction with no accuracy drop and negligible hardware overhead.

📄 PDF Abstract BibTeX arXiv:2210.17047

Code (0)

등록된 구현이 없습니다.

Tasks

QuantizationSensitivity

Similar Papers 제목 키워드 기반

Ax-BxP: Approximate Blocked Computation for Precision-Reconfigurable Deep Neural Network Acceleration

2020-11-25 · Reena Elangovan, Shubham Jain, Anand Raghunathan

Precision scaling has emerged as a popular technique to optimize the compute and storage requirements of Deep Neural Networks (DNNs). Efforts toward creating ultra-low-precision (sub-8-bit) DNNs suggest that the minimum …

Sortblock: Similarity-Aware Feature Reuse for Diffusion Model

2025-08-01 · Hanqi Chen, Xu Zhang, Xiaoliu Guan, Lielin Jiang 외 arxiv

Diffusion Transformers (DiTs) have demonstrated remarkable generative capabilities, particularly benefiting from Transformer architectures that enhance visual and artistic fidelity. However, their inherently sequential d…

Pipe-BD: Pipelined Parallel Blockwise Distillation

2023-01-29 · Hongsun Jang, Jaewon Jung, Jaeyong Song, Joonsang Yu 외

Training large deep neural network models is highly challenging due to their tremendous computational and memory requirements. Blockwise distillation provides one promising method towards faster convergence by splitting …

GPU

CorGi: Contribution-Guided Block-Wise Interval Caching for Training-Free Acceleration of Diffusion Transformers

2025-12-30 · Yonglak Son, Suhyeok Kim, Seungryong Kim, Young Geun Kim arxiv

Diffusion transformer (DiT) achieves remarkable performance in visual generation, but its iterative denoising process combined with larger capacity leads to a high inference cost. Recent works have demonstrated that the …

Dynamically Reconfigurable Variable-precision Sparse-Dense Matrix Acceleration in Tensorflow Lite

2023-04-17 · Jose Nunez-Yanez, Andres Otero, Eduardo de la Torre

In this paper, we present a dynamically reconfigurable hardware accelerator called FADES (Fused Architecture for DEnse and Sparse matrices). The FADES design offers multiple configuration options that trade off paralleli…