paper-with-me

Papers

DSV: Exploiting Dynamic Sparsity to Accelerate Large-Scale Video DiT Training

2025-02-11 · Xin Tan, Yuetao Chen, Yimin Jiang, Xing Chen, Kun Yan, Nan Duan, Yibo Zhu, Daxin Jiang, Hong Xu

Diffusion Transformers (DiTs) have shown remarkable performance in generating high-quality videos. However, the quadratic complexity of 3D full attention remains a bottleneck in scaling DiT training, especially with high-definition, lengthy videos, where it can consume up to 95% of processing time and demand specialized context parallelism. This paper introduces DSV to accelerate video DiT training by leveraging the dynamic attention sparsity we empirically observe. DSV uses a two-stage algorithm to capture the dynamic sparsity patterns via low-rank based approximation of the original query and key. It employs custom kernels to efficiently identify critical key-value pairs and compute the sparse attention. To accommodate the new sparsity dimension, DSV adopts a hybrid sparsity-aware context parallelism that re-balances the skewed workload across attention heads and blocks due to sparsity heterogeneity. DSV achieves up to 3.02x higher training throughput, scaling to 128 GPUs and 520k token lengths, without quality loss.

📄 PDF Abstract BibTeX arXiv:2502.07590

Code (0)

등록된 구현이 없습니다.

Methods 이 논문이 사용한 방법론

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 $x$…
Attention 설명 없음

Similar Papers 제목 키워드 기반

Exploiting On-chip Heterogeneity of Versal Architecture for GNN Inference Acceleration

2023-08-04 · Paul Chen, Pavan Manjunath, Sasindu Wijeratne, Bingyi Zhang 외

Graph Neural Networks (GNNs) have revolutionized many Machine Learning (ML) applications, such as social network analysis, bioinformatics, etc. GNN inference can be accelerated by exploiting data sparsity in the input gr…

CPUGPU

S2TA: Exploiting Structured Sparsity for Energy-Efficient Mobile CNN Acceleration

2021-07-16 · Zhi-Gang Liu, Paul N. Whatmough, Yuhao Zhu, Matthew Mattina

Exploiting sparsity is a key technique in accelerating quantized convolutional neural network (CNN) inference on mobile devices. Prior sparse CNN accelerators largely exploit un-structured sparsity and achieve significan…

Speeding up Convolutional Neural Networks By Exploiting the Sparsity of Rectifier Units

2017-04-25 · Shaohuai Shi, Xiaowen Chu

Rectifier neuron units (ReLUs) have been widely used in deep convolutional networks. An ReLU converts negative values to zeros, and does not change positive values, which leads to a high sparsity of neurons. In this work…

Exposing and Exploiting Fine-Grained Block Structures for Fast and Accurate Sparse Training

2022-11-01 · NIPS 2022 11 · Peng Jiang, Lihan Hu, Shihui Song

Sparse training is a popular technique to reduce the overhead of training large models. Although previous work has shown promising results for nonstructured sparse models, it is still unclear whether a sparse model with …

SparseTrain: Exploiting Dataflow Sparsity for Efficient Convolutional Neural Networks Training

2020-07-21 · Pengcheng Dai, Jianlei Yang, Xucheng Ye, Xingzhou Cheng 외

Training Convolutional Neural Networks (CNNs) usually requires a large number of computational resources. In this paper, \textit{SparseTrain} is proposed to accelerate CNN training by fully exploiting the sparsity. It ma…