Runtime Performance Prediction for Deep Learning Models with Graph Neural Network

Gao, Yanjie; Gu, Xianyu; Zhang, Hongyu; Lin, Haoxiang; Yang, Mao

doi:10.1109/icse-seip58684.2023.00039

Cited by 24 publications

(1 citation statement)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Among the most important features considered are the batch size, dataset size, and the pre-trained DNN utilized. The runtime prediction of DNN is also addressed in [15] but focused on GPU architectures. Among the input features considered, the most relevant were the GPU model and the selected DNN architecture.…”

Section: Related Workmentioning

confidence: 99%

Modeling ML Efficiency in Telecommunications

Corona,

Teixeira,

Antunes

et al. 2024

Preprint

View full text Add to dashboard Cite

Machine learning (ML) models are envisioned to play a fundamental role in Beyond 5G networks. However, training these models presents significant challenges due to the vast amounts of data involved and the complexity of the algorithms, requiring substantial training time and computing resources. This study offers a comprehensive analysis of how dataset characteristics, model complexity, and the number of iterations influence model performance and training time. Furthermore, we take pioneering steps towards modeling the performance and training time of ML models, specifically within the telecommunications sector. Our findings can pave the way for informed decision-making processes in telecommunications, enabling a comprehensive understanding of expected performance and training time requirements. Leading to reduced development times and optimized energy consumption and resource allocation.

show abstract

Section: Related Workmentioning

confidence: 99%

Modeling ML Efficiency in Telecommunications

Corona,

Teixeira,

Antunes

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

A lightweight performance proxy for deep‐learning model training on Amazon SageMaker

Keller Tesser,

Marques,

Borin

2024

Concurrency and Computation

View full text Add to dashboard Cite

SummaryCloud computing has become popular for training deep‐learning (DL) models, avoiding the costs of acquiring and maintaining on‐premise systems. SageMaker is a cloud service that automates the execution of DL workloads. Its features include automatic hyperparameter optimization and use of spot instances. Nonetheless, it does not assist in selecting the right instance type for a workload. In public clouds, rent price depends on the configuration of the chosen instance type. Advanced and faster instances are typically more expensive, but not always the best choice. To select the optimal instance type, users must compare the workload's relative performance (and hence cost) on several candidates. Building on the execution profiles of multiple DL applications, we model the performance and cost of training DL applications on SageMaker and propose a lightweight technique to estimate these at low temporal and monetary cost. This method is a performance proxy that can be used to replace more expensive performance measurement procedures. So, it could speed up any technique that relies on such measurements. We show how it can help cloud customers seeking suitable instance types to train DL models, and that it can accurately predict the performance of different instance types when training these models on SageMaker.

show abstract