CAEFL: composable and environment aware federated learning models

Xu, Ruomeng; Michala, Anna Lito; Trinder, Phil

doi:10.1145/3546186.3549927

Cited by 3 publications

(1 citation statement)

References 29 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…No contexto da linguagem Elixir, esta é usada por [Xu et al 2022] para a realizac ¸ão do treinamento federado de redes neurais e, diferente do trabalho anterior, o treinamento nesse trabalho foi realizado numa máquina com grande capacidade de hardware dedicado. Nesse trabalho, o paralelismo local de cada nodo também não é explorado.…”

Section: Trabalhos Relacionadosunclassified

Treinamento Paralelo de Redes Neurais usando o Modelo de Atores

Bartsch¹,

Bois²

2023

Anais Da XXIII Escola Regional De Alto Desempenho Da Região Sul (ERAD-RS 2023)

View full text Add to dashboard Cite

O Modelo de Atores usa concorrência baseada em processos que não compartilham memória e que se comunicam por troca de mensagens, e tem sido usado no treinamento federado de redes neurais. Esse artigo é um trabalho em andamento que investiga o treinamento paralelo de redes neurais em dispositivos multiprocessados, utilizando o Modelo de Atores para distribuir e processar batches de treinamento, com objetivo de melhorar algoritmos de treinamento.

show abstract

Section: Trabalhos Relacionadosunclassified

Treinamento Paralelo de Redes Neurais usando o Modelo de Atores

Bartsch¹,

Bois²

2023

Anais Da XXIII Escola Regional De Alto Desempenho Da Região Sul (ERAD-RS 2023)

View full text Add to dashboard Cite

show abstract

GPotion: An embedded DSL for GPU programming in Elixir

Du Bois,

Cavalheiro

2023

Proceedings of the XXVII Brazilian Symposium on Programming Languages

View full text Add to dashboard Cite

Hok: Higher-Order GPU kernels in Elixir

Du Bois,

Perlin,

Antunes

et al. 2024

Anais Do XXVIII Simpósio Brasileiro De Linguagens De Programação (SBLP 2024)

View full text Add to dashboard Cite

GPUs (Graphics Processing Units) are usually programmed using low-level languages like CUDA or OpenCL. Although these languages allow the implementation of very optimized software, they are difficult to program due to their low-level nature, where programmers have to mix coordination code, i.e., how tasks are created and distributed, with the actual computation code. In this paper we present Hok, an extension to the Elixir functional language that allows the implementation of higher-order GPU kernels, granting programmers the ability to clearly separate coordination from computation. The Hok system provides a DSL (Domain-Specific Language) for writing low-level GPU kernels that can be parameterized with the computation code. Hok allows device functions, including anonymous functions, to be created and referenced in the host code so that they can configure a kernel before it is launched.We demonstrate that Hok can be used to implement high-level abstractions such as algorithmic skeletons and array comprehensions. We also present experiments that demonstrate the usability of the current implementation of Hok, and show that high speedups can be obtained in comparison to pure Elixir, specially in computationally intensive programs with large inputs.

show abstract

CAEFL: composable and environment aware federated learning models

Abstract: There may be differences between this version and the published version. You are advised to consult the publisher's version if you wish to cite from it.

Cited by 3 publications

References 29 publications

Treinamento Paralelo de Redes Neurais usando o Modelo de Atores

Treinamento Paralelo de Redes Neurais usando o Modelo de Atores

GPotion: An embedded DSL for GPU programming in Elixir

Hok: Higher-Order GPU kernels in Elixir

Contact Info

Product

Resources

About