paper-with-me

Papers

OmniLearn: A Framework for Distributed Deep Learning over Heterogeneous Clusters

2025-03-21 · Sahil Tyagi, Prateek Sharma

Deep learning systems are optimized for clusters with homogeneous resources. However, heterogeneity is prevalent in computing infrastructure across edge, cloud and HPC. When training neural networks using stochastic gradient descent techniques on heterogeneous resources, performance degrades due to stragglers and stale updates. In this work, we develop an adaptive batch-scaling framework called OmniLearn to mitigate the effects of heterogeneity in distributed training. Our approach is inspired by proportional controllers to balance computation across heterogeneous servers, and works under varying resource availability. By dynamically adjusting worker mini-batches at runtime, OmniLearn reduces training time by 14-85%. We also investigate asynchronous training, where our techniques improve accuracy by up to 6.9%.

📄 PDF Abstract BibTeX arXiv:2503.17469

Code (1)

sahiltyagi4/OmniLearn 공식 구현 pytorch

Tasks

Deep Learning

Similar Papers 제목 키워드 기반

VoTranhOmniLearner - A Self-Evolving Digital Universe for Simulating Consciousness and Societal Dynamics

2025-04-26 · Independent publication 2025 4 · Vi Nhat Son

VoTranhOmniLearner is an innovative simulation framework designed to model a dynamic digital universe with thousands of entities, intricate social interactions, and self-reflective processes. By integrating machine learn…

Management

Holmes: Towards Distributed Training Across Clusters with Heterogeneous NIC Environment

2023-12-06 · Fei Yang, Shuang Peng, Ning Sun, Fangyu Wang 외

Large language models (LLMs) such as GPT-3, OPT, and LLaMA have demonstrated remarkable accuracy in a wide range of tasks. However, training these models can incur significant expenses, often requiring tens of thousands …

GPUScheduling

ABS-SGD: A Delayed Synchronous Stochastic Gradient Descent Algorithm with Adaptive Batch Size for Heterogeneous GPU Clusters

2023-08-29 · Xin Zhou, Ling Chen, Houming Wu

As the size of models and datasets grows, it has become increasingly common to train models in parallel. However, existing distributed stochastic gradient descent (SGD) algorithms suffer from insufficient utilization of …

GPU

Helix: Serving Large Language Models over Heterogeneous GPUs and Network via Max-Flow

2024-06-03 · Yixuan Mei, Yonghao Zhuang, Xupeng Miao, Juncheng Yang 외

This paper introduces Helix, a distributed system for high-throughput, low-latency large language model (LLM) serving in heterogeneous GPU clusters. The key idea behind Helix is to formulate inference computation of LLMs…

GPULanguage ModelingLanguage ModellingLarge Language Model+1

Taming Resource Heterogeneity In Distributed ML Training With Dynamic Batching

2023-05-20 · Sahil Tyagi, Prateek Sharma

Current techniques and systems for distributed model training mostly assume that clusters are comprised of homogeneous servers with a constant resource availability. However, cluster heterogeneity is pervasive in computi…

CPUGPU