中文
相关论文

相关论文: AMDP: Asynchronous Multi-Directional Pipeline Para…

200 篇论文

Particle tracking has several important applications for solute transport studies in aquifer systems. Travel time distribution at observation points, particle coordinates in time and streamlines are some practical results providing…

分布式、并行与集群计算 · 计算机科学 2023-02-28 Rodrigo Pérez-Illanes , Daniel Fernàndez-Garcia

Many real-world optimization models contain exploitable sparsity and block structure, but this structure is often obscured in algebraic form, limiting the effectiveness of modern parallel algorithms. We propose an automatic pipeline that…

最优化与控制 · 数学 2026-03-23 Kaizhao Sun , Baihao Wu , Kun Yuan , Wotao Yin

Current structural pruning methods face two significant limitations: (i) they often limit pruning to finer-grained levels like channels, making aggressive parameter reduction challenging, and (ii) they focus heavily on parameter and FLOP…

计算机视觉与模式识别 · 计算机科学 2025-04-04 Xinglong Sun , Barath Lakshmanan , Maying Shen , Shiyi Lan , Jingde Chen , Jose M. Alvarez

Multi-step LLM pipelines can solve complex tasks, but jointly optimizing prompts across steps remains challenging due to missing step-level supervision and inter-step dependency. We propose ADOPT, an adaptive dependency-guided joint prompt…

计算与语言 · 计算机科学 2026-04-09 Minjun Zhao , Xinyu Zhang , Shuai Zhang , Deyang Li , Ruifeng Shi

In recent years, Large Language Models (LLMs) have exhibited remarkable capabilities, driving advancements in real-world applications. However, training LLMs on increasingly long input sequences imposes significant challenges due to high…

分布式、并行与集群计算 · 计算机科学 2025-03-14 Qiaoling Chen , Shenggui Li , Wei Gao , Peng Sun , Yonggang Wen , Tianwei Zhang

Pipeline parallelism (PP) is widely used for training large language models (LLMs), yet its scalability is often constrained by high activation memory consumption as the number of in-flight microbatches grows with the degree of PP. In this…

机器学习 · 计算机科学 2025-07-01 Xinyi Wan , Penghui Qi , Guangxing Huang , Min Lin , Jialin Li

Semi-dense feature matching methods have shown strong performance in challenging scenarios. However, the existing pipeline relies on a global search across the entire feature map to establish coarse matches, limiting further improvements in…

计算机视觉与模式识别 · 计算机科学 2025-08-04 Peiqi Chen , Lei Yu , Yi Wan , Yingying Pei , Xinyi Liu , Yongxiang Yao , Yingying Zhang , Lixiang Ru , Liheng Zhong , Jingdong Chen , Ming Yang , Yongjun Zhang

We design and implement a ready-to-use library in PyTorch for performing micro-batch pipeline parallelism with checkpointing proposed by GPipe (Huang et al., 2019). In particular, we develop a set of design components to enable…

分布式、并行与集群计算 · 计算机科学 2020-04-22 Chiheon Kim , Heungsub Lee , Myungryong Jeong , Woonhyuk Baek , Boogeon Yoon , Ildoo Kim , Sungbin Lim , Sungwoong Kim

Scaling multi-dimensional transformers to long sequences is indispensable across various domains. However, the challenges of large memory requirements and slow speeds of such sequences necessitate sequence parallelism. All existing…

分布式、并行与集群计算 · 计算机科学 2025-05-13 Xuanlei Zhao , Shenggan Cheng , Chang Chen , Zangwei Zheng , Ziming Liu , Zheming Yang , Yang You

In this paper we analyze, evaluate, and improve the performance of training generalized linear models on modern CPUs. We start with a state-of-the-art asynchronous parallel training algorithm, identify system-level performance bottlenecks,…

机器学习 · 计算机科学 2018-12-20 Nikolas Ioannou , Celestine Dünner , Kornilios Kourtis , Thomas Parnell

The increasing demand for intelligent mobile applications has made multi-agent collaboration with Transformer-based large language models (LLMs) essential in mobile edge computing (MEC) networks. However, training LLMs in such environments…

系统与控制 · 电气工程与系统科学 2025-09-25 Jiewei Chen , Xiumei Deng , Zehui Xiong , Shaoyong Guo , Xuesong Qiu , Ping Wang , Dusit Niyato

Pipeline parallelism is an essential technique in the training of large-scale Transformer models. However, it suffers from imbalanced memory consumption, leading to insufficient memory utilization. The BPipe technique was proposed to…

机器学习 · 计算机科学 2024-01-05 Mincong Huang , Chao Wang , Chi Ma , Yineng Zhang , Peng Zhang , Lei Yu

Currently, training large-scale deep learning models is typically achieved through parallel training across multiple GPUs. However, due to the inherent communication overhead and synchronization delays in traditional model parallelism…

计算机视觉与模式识别 · 计算机科学 2024-11-21 Xiuyuan Guo , Chengqi Xu , Guinan Guo , Feiyu Zhu , Changpeng Cai , Peizhe Wang , Xiaoming Wei , Junhao Su , Jialin Gao

On-device Deep Neural Network (DNN) training has been recognized as crucial for privacy-preserving machine learning at the edge. However, the intensive training workload and limited onboard computing resources pose significant challenges to…

分布式、并行与集群计算 · 计算机科学 2024-08-16 Shengyuan Ye , Liekang Zeng , Xiaowen Chu , Guoliang Xing , Xu Chen

Aiming at solving large-scale learning problems, this paper studies distributed optimization methods based on the alternating direction method of multipliers (ADMM). By formulating the learning problem as a consensus problem, the ADMM can…

分布式、并行与集群计算 · 计算机科学 2016-05-04 Tsung-Hui Chang , Mingyi Hong , Wei-Cheng Liao , Xiangfeng Wang

In commercial systems, a pervasive requirement for automatic data preparation (ADP) is to transfer relational data from disparate sources to targets with standardized schema specifications. Previous methods rely on labor-intensive…

人工智能 · 计算机科学 2025-09-23 Congcong Ge , Yachuan Liu , Yixuan Tang , Yifan Zhu , Yaofeng Tu , Yunjun Gao

The increasing size of deep learning models has made distributed training across multiple devices essential. However, current methods such as distributed data-parallel training suffer from large communication and synchronization overheads…

机器学习 · 计算机科学 2025-02-10 Cabrel Teguemne Fokam , Khaleelulla Khan Nazeer , Lukas König , David Kappel , Anand Subramoney

Training large models with distributed data parallelism (DDP) requires frequent communication of gradients across workers, which can saturate bandwidth. Infrequent communication strategies (e.g., Local SGD) reduce this overhead but, when…

The alternating direction method of multipliers (ADMM) has been recognized as a versatile approach for solving modern large-scale machine learning and signal processing problems efficiently. When the data size and/or the problem dimension…

分布式、并行与集群计算 · 计算机科学 2016-05-04 Tsung-Hui Chang , Wei-Cheng Liao , Mingyi Hong , Xiangfeng Wang

Parallelism is a ubiquitous method for accelerating machine learning algorithms. However, theoretical analysis of parallel learning is usually done in an algorithm- and protocol-specific setting, giving little insight about how changes in…

机器学习 · 计算机科学 2020-06-09 Yucheng Lu , Jack Nash , Christopher De Sa