English
Related papers

Related papers: MegaTrain: Full Precision Training of 100B+ Parame…

200 papers

Biosignals exhibit substantial cross-subject and cross-session variability, inducing severe domain shifts that degrade post-deployment performance for small, edge-oriented AI models. On-device adaptation is therefore essential to both…

Machine Learning · Computer Science 2026-04-24 Run Wang , Victor J. B. Jung , Philip Wiese , Sebastian Frey , Giusy Spacone , Francesco Conti , Alessio Burrello , Luca Benini

As deep learning models become increasingly large, they pose significant challenges in heterogeneous devices environments. The size of deep learning models makes it difficult to deploy them on low-power or resource-constrained devices,…

Machine Learning · Computer Science 2023-11-27 Mert Unsal , Ali Maatouk , Antonio De Domenico , Nicola Piovesan , Fadhel Ayed

Training large transformer models is one of the most important computational challenges of modern AI. In this paper, we show how to significantly accelerate training of large transformer models by reducing activation recomputation.…

Machine Learning · Computer Science 2022-05-12 Vijay Korthikanti , Jared Casper , Sangkug Lym , Lawrence McAfee , Michael Andersch , Mohammad Shoeybi , Bryan Catanzaro

Training on the Edge enables neural networks to learn continuously from new data after deployment on memory-constrained edge devices. Previous work is mostly concerned with reducing the number of model parameters which is only beneficial…

Machine Learning · Computer Science 2021-11-01 Abdelrahman Hosny , Marina Neseem , Sherief Reda

Typically, Ultra-deep neural network(UDNN) tends to yield high-quality model, but its training process is usually resource intensive and time-consuming. Modern GPU's scarce DRAM capacity is the primary bottleneck that hinders the…

Machine Learning · Computer Science 2019-06-21 Jinrong Guo , Wantao Liu , Wang Wang , Qu Lu , Songlin Hu , Jizhong Han , Ruixuan Li

Efficiently training LLMs with long sequences is important yet challenged by the massive computation and memory requirements. Sequence parallelism has been proposed to tackle these problems, but existing methods suffer from scalability or…

Distributed, Parallel, and Cluster Computing · Computer Science 2024-06-27 Diandian Gu , Peng Sun , Qinghao Hu , Ting Huang , Xun Chen , Yingtong Xiong , Guoteng Wang , Qiaoling Chen , Shangchun Zhao , Jiarui Fang , Yonggang Wen , Tianwei Zhang , Xin Jin , Xuanzhe Liu

Despite the dominance and effectiveness of scaling, resulting in large networks with hundreds of billions of parameters, the necessity to train overparameterized models remains poorly understood, while training costs grow exponentially. In…

Computation and Language · Computer Science 2023-12-12 Vladislav Lialin , Namrata Shivagunde , Sherin Muckatira , Anna Rumshisky

Videos are big, complex to pre-process, and slow to train on. State-of-the-art large-scale video models are trained on clusters of 32 or more GPUs for several days. As a consequence, academia largely ceded the training of large video models…

Computer Vision and Pattern Recognition · Computer Science 2023-09-29 Yue Zhao , Philipp Krähenbühl

The rapid growth in machine learning models, especially in natural language processing and computer vision, has led to challenges when running these models on hardware with limited resources. This paper introduces Superpipeline, a new…

Machine Learning · Computer Science 2024-10-14 Reza Abbasi , Sernam Lim

In machine translation tasks, the relationship between model complexity and performance is often presumed to be linear, driving an increase in the number of parameters and consequent demands for computational resources like multiple GPUs.…

Computation and Language · Computer Science 2023-08-14 Luv Verma , Ketaki N. Kolhatkar

Large language models (LLMs) are computationally intensive. The computation workload and the memory footprint grow quadratically with the dimension (layer width). Most of LLMs' parameters come from the linear layers of the transformer…

Machine Learning · Computer Science 2024-02-22 Xiao-Yang Liu , Jie Zhang , Guoxuan Wang , Weiqing Tong , Anwar Walid

GPUs have been favored for training deep learning models due to their highly parallelized architecture. As a result, most studies on training optimization focus on GPUs. There is often a trade-off, however, between cost and efficiency when…

Multimodal large language models (LLMs) empower LLMs to ingest inputs and generate outputs in multiple forms, such as text, image, and audio. However, the integration of multiple modalities introduces heterogeneity in both the model and…

Distributed, Parallel, and Cluster Computing · Computer Science 2025-12-01 Zili Zhang , Yinmin Zhong , Yimin Jiang , Hanpeng Hu , Jianjian Sun , Zheng Ge , Yibo Zhu , Daxin Jiang , Xin Jin

Training foundation models, such as GPT-3 and PaLM, can be extremely expensive, often involving tens of thousands of GPUs running continuously for months. These models are typically trained in specialized clusters featuring fast,…

Distributed, Parallel, and Cluster Computing · Computer Science 2023-06-22 Binhang Yuan , Yongjun He , Jared Quincy Davis , Tianyi Zhang , Tri Dao , Beidi Chen , Percy Liang , Christopher Re , Ce Zhang

Recommendation models rely on deep learning networks and large embedding tables, resulting in computationally and memory-intensive processes. These models are typically trained using hybrid CPU-GPU or GPU-only configurations. The hybrid…

Hardware Architecture · Computer Science 2024-04-30 Muhammad Adnan , Yassaman Ebrahimzadeh Maboud , Divya Mahajan , Prashant J. Nair

The rapid growth of large language models (LLMs) has outpaced the evolution of single-GPU hardware, making model scale increasingly constrained by memory capacity rather than computation. While modern training systems extend GPU memory…

Operating Systems · Computer Science 2026-04-08 Zhengqing Yuan , Lichao Sun , Yanfang Ye

Because of the superior feature representation ability of deep learning, various deep Click-Through Rate (CTR) models are deployed in the commercial systems by industrial companies. To achieve better performance, it is necessary to train…

Information Retrieval · Computer Science 2021-05-12 Huifeng Guo , Wei Guo , Yong Gao , Ruiming Tang , Xiuqiang He , Wenzhi Liu

Existing general purpose frameworks for gigantic model training, i.e., dense models with billions of parameters, cannot scale efficiently on cloud environment with various networking conditions due to large communication overheads. In this…

Distributed, Parallel, and Cluster Computing · Computer Science 2022-10-31 Zhen Zhang , Shuai Zheng , Yida Wang , Justin Chiu , George Karypis , Trishul Chilimbi , Mu Li , Xin Jin

Spatio-temporal modeling as a canonical task of multivariate time series forecasting has been a significant research topic in AI community. To address the underlying heterogeneity and non-stationarity implied in the graph streams, in this…

Large-scale model training has been a playing ground for a limited few requiring complex model refactoring and access to prohibitively expensive GPU clusters. ZeRO-Offload changes the large model training landscape by making large model…

Distributed, Parallel, and Cluster Computing · Computer Science 2021-01-19 Jie Ren , Samyam Rajbhandari , Reza Yazdani Aminabadi , Olatunji Ruwase , Shuangyan Yang , Minjia Zhang , Dong Li , Yuxiong He