Extra-Deep

This repository contains the source code for Extra-Deep as described in the paper "Extra-Deep: Automated Empirical Performance Modeling for Distributed Deep Learning". Furthermore, it contains the source code for the application benchmarks and some sample performance measuements to get started with an analysis.

Please cite the dataset in your publications if it helps your research:

@dataset{ritter_2024_14183393,
    author       = {Ritter, Marcus},
    title        = {{Performance Measurement Dataset Example for Extra- 
                    Deep CIFAR-10}},
    month        = nov,
    year         = 2024,
    publisher    = {Zenodo},
    version      = {1.0},
    doi          = {10.5281/zenodo.14183393},
    url          = {https://doi.org/10.5281/zenodo.14183393}
}

Please cite the paper in your publications if it helps your research:

@INPROCEEDINGS{ritter_ea:protools:2023,
    author = {Ritter, Marcus and Wolf, Felix},
    month = nov,
    title = {Extra-Deep: Automated Empirical Performance Modeling for Distributed Deep Learning},
    booktitle = {Proc. of the Workshop on Programming and Performance Visualization Tools (ProTools), held in conjunction with the Supercomputing Conference (SC23), Denver, CO, USA},
    year = {2023},
    pages = {1345--1356},
    publisher = {ACM},
    isbn = {9798400707858},
    doi = {10.1145/3624062.3624204}
}