English
Related papers

Related papers: OptPipe: Memory- and Scheduling-Optimized Pipeline…

200 papers

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…

Machine Learning · Computer Science 2024-01-05 Mincong Huang , Chao Wang , Chi Ma , Yineng Zhang , Peng Zhang , Lei Yu

Fine-tuning Large Language Models (LLMs) on consumer-grade GPUs is highly cost-effective, yet constrained by limited GPU memory and slow PCIe interconnects. Pipeline parallelism combined with CPU offloading mitigates these hardware…

Distributed, Parallel, and Cluster Computing · Computer Science 2026-05-01 Yibin Luo , Shiwei Gao , Huichuan Zheng , Youyou Lu , Jiwu Shu

We introduce Breadth-First Pipeline Parallelism, a novel training schedule which optimizes the combination of pipeline and data parallelism. Breadth-First Pipeline Parallelism lowers training time, cost and memory usage by combining a high…

Distributed, Parallel, and Cluster Computing · Computer Science 2023-07-10 Joel Lamy-Poirier

To support large-scale model training, split learning (SL) enables multiple edge devices/servers to share the intensive training workload. However, most existing works on SL focus solely on two-tier model splitting. Moreover, while some…

Networking and Internet Architecture · Computer Science 2025-09-19 Wei Wei , Zheng Lin , Tao Li , Xuanheng Li , Xianhao Chen

Training large language models (LLMs) is known to be challenging because of the huge computational and memory capacity requirements. To address these issues, it is common to use a cluster of GPUs with 3D parallelism, which splits a model…

Distributed, Parallel, and Cluster Computing · Computer Science 2024-05-29 Jinkyu Yim , Jaeyong Song , Yerim Choi , Jaebeen Lee , Jaewon Jung , Hongsun Jang , Jinho Lee

We present JaxPP, a system for efficiently scaling the training of large deep learning models with flexible pipeline parallelism. We introduce a seamless programming model that allows implementing user-defined pipeline schedules for…

Distributed, Parallel, and Cluster Computing · Computer Science 2024-12-20 Anxhelo Xhebraj , Sean Lee , Hanfeng Chen , Vinod Grover

Large language models have led to state-of-the-art accuracies across a range of tasks. However, training these models efficiently is challenging for two reasons: a) GPU memory capacity is limited, making it impossible to fit large models on…

Data and pipeline parallelism are ubiquitous for training of Large Language Models (LLM) on distributed nodes. Driven by the need for cost-effective training, recent work explores efficient communication arrangement for end to end training.…

Machine Learning · Computer Science 2025-02-28 Nikolay Blagoev , Lydia Yiyu Chen , Oğuzhan Ersoy

The emergence of large language models (LLMs) relies heavily on distributed training strategies, among which pipeline parallelism plays a crucial role. As LLMs' training sequence length extends to 32k or even 128k, the current pipeline…

Distributed, Parallel, and Cluster Computing · Computer Science 2024-11-12 Ao Sun , Weilin Zhao , Xu Han , Cheng Yang , Xinrong Zhang , Zhiyuan Liu , Chuan Shi , Maosong Sun

The occurrence of bubbles in pipeline parallelism is an inherent limitation that can account for more than 40% of the large language model (LLM) training time and is one of the main reasons for the underutilization of GPU resources in LLM…

Distributed, Parallel, and Cluster Computing · Computer Science 2025-04-29 Jiashu Zhang , Zihan Pan , Molly , Xu , Khuzaima Daudjee , Sihang Liu

Training large language models (LLMs) is fundamentally constrained by limited device memory and costly inter-device communication. Although pipeline parallelism alleviates memory pressure by partitioning models across devices, it incurs…

Machine Learning · Computer Science 2025-11-14 Houming Wu , Ling Chen

Many state-of-the-art ML results have been obtained by scaling up the number of parameters in existing models. However, parameters and activations for such large models often do not fit in the memory of a single accelerator device; this…

Machine Learning · Computer Science 2021-07-23 Deepak Narayanan , Amar Phanishayee , Kaiyu Shi , Xie Chen , Matei Zaharia

Training Deep Neural Networks (DNNs) with billions of parameters generally involves pipeline-parallel (PP) execution. Unfortunately, PP model training can use GPUs inefficiently, especially at large scale, due to idle GPU time caused by…

Distributed, Parallel, and Cluster Computing · Computer Science 2024-10-11 Daiyaan Arfeen , Zhen Zhang , Xinwei Fu , Gregory R. Ganger , Yida Wang

Transformers and LLMs have seen rapid adoption in all domains. Their sizes have exploded to hundreds of billions of parameters and keep increasing. Under these circumstances, the training of transformers is slow and often takes in the order…

Distributed, Parallel, and Cluster Computing · Computer Science 2024-06-18 Avinash Maurya , Jie Ye , M. Mustafa Rafique , Franck Cappello , Bogdan Nicolae

Large-scale language models have become increasingly challenging and expensive to train. Among various methods addressing this issue, Pipeline Parallelism has been widely employed to accommodate massive model weights within limited GPU…

Distributed, Parallel, and Cluster Computing · Computer Science 2023-08-31 Ziming Liu , Shenggan Cheng , Haotian Zhou , Yang You

In training of modern large natural language processing (NLP) models, it has become a common practice to split models using 3D parallelism to multiple GPUs. Such technique, however, suffers from a high overhead of inter-node communication.…

Machine Learning · Computer Science 2023-01-25 Jaeyong Song , Jinkyu Yim , Jaewon Jung , Hongsun Jang , Hyung-Jin Kim , Youngsok Kim , Jinho Lee

The rapid evolution of large language models (LLMs) has made geographically distributed training necessary due to GPU scarcity within a single cloud region. In such cross-region settings, Pipeline Parallelism (PP) is…

Distributed, Parallel, and Cluster Computing · Computer Science 2026-05-26 Han Zhang , Jianchun Liu , Hongli Xu

Training LLMs larger than the aggregated memory of multiple GPUs is increasingly necessary due to the faster growth of LLM sizes compared to GPU memory. To this end, multi-tier host memory or disk offloading techniques are proposed by state…

Distributed, Parallel, and Cluster Computing · Computer Science 2025-09-03 Avinash Maurya , M. Mustafa Rafique , Franck Cappello , Bogdan Nicolae

Large model training often uses recomputation to alleviate memory pressure and pipelines to exploit the parallelism of data, tensors, and devices. However, existing recomputation approaches may incur high overhead when training real-world…

Distributed, Parallel, and Cluster Computing · Computer Science 2025-03-31 Ping Chen , Wenjie Zhang , Shuibing He , Weijian Chen , Siling Yang , Kexin Huang , Yanlong Yin , Xuan Zhan , Yingjie Gu , Zhuwei Peng , Yi Zheng , Zhefeng Wang , Gang Chen

Pipeline parallelism enables efficient training of Large Language Models (LLMs) on large-scale distributed accelerator clusters. Yet, pipeline bubbles during startup and tear-down reduce the utilization of accelerators. Although efficient…

Machine Learning · Computer Science 2023-05-16 Kazuki Osawa , Shigang Li , Torsten Hoefler