BUTTER-E - Energy Consumption Data for the BUTTER Empirical Deep Learning Dataset

Tripp, Charles; Perr-Sauer, Jordan; Bensen, Erik; Gafur, Jamil; Nag, Ambarish; Purkayastha, Avi

doi:10.25984/2329316

Title: BUTTER-E - Energy Consumption Data for the BUTTER Empirical Deep Learning Dataset

Dataset
Other Related Research

Abstract

The BUTTER-E - Energy Consumption Data for the BUTTER Empirical Deep Learning Dataset adds node-level energy consumption data from watt-meters to the primary sweep of the BUTTER - Empirical Deep Learning Dataset. This dataset contains energy consumption and performance data from 63,527 individual experimental runs spanning 30,582 distinct configurations: 13 datasets, 20 sizes (number of trainable parameters), 8 network "shapes", and 14 depths on both CPU and GPU hardware collected using node-level watt-meters. This dataset reveals the complex relationship between dataset size, network structure, and energy use, and highlights the impact of cache effects. BUTTER-E is intended to be joined with the BUTTER dataset (see "BUTTER - Empirical Deep Learning Dataset on OEDI" resource below) which characterizes the performance of 483k distinct fully connected neural networks but does not include energy measurements.

Authors:

Tripp, Charles; Perr-Sauer, Jordan; Bensen, Erik; Gafur, Jamil; Nag, Ambarish; Purkayastha, Avi

National Renewable Energy Laboratory

Publication Date:: Fri Dec 30 04:00:00 UTC 2022

Other Number(s):: 5991

Research Org.:: DOE Open Energy Data Initiative (OEDI); National Renewable Energy Laboratory

Sponsoring Org.:: USDOE Office of Science (SC), Advanced Scientific Computing Research (SC-31)

Collaborations:: National Renewable Energy Laboratory

Subject:: Array; BUTTER; BUTTER-E; benchmark; computational science; deep learning; efficient; empirical deep learning; empirical machine learning; energy; energy consumption; energy efficiency; energy use; green computing; machine learning; model; network structure; neural networks; node-level; power; power consumption; training; training efficiency

OSTI Identifier:: 2329316

DOI:: https://doi.org/10.25984/2329316

Citation Formats


                    Tripp, Charles, Perr-Sauer, Jordan, Bensen, Erik, Gafur, Jamil, Nag, Ambarish, and Purkayastha, Avi. BUTTER-E - Energy Consumption Data for the BUTTER Empirical Deep Learning Dataset.  United States: N. p., 2022. 
        Web.  doi:10.25984/2329316.

Copy to clipboard


                    Tripp, Charles, Perr-Sauer, Jordan, Bensen, Erik, Gafur, Jamil, Nag, Ambarish, & Purkayastha, Avi. BUTTER-E - Energy Consumption Data for the BUTTER Empirical Deep Learning Dataset.  United States.  doi:https://doi.org/10.25984/2329316

Copy to clipboard


                    Tripp, Charles, Perr-Sauer, Jordan, Bensen, Erik, Gafur, Jamil, Nag, Ambarish, and Purkayastha, Avi. 2022.  
"BUTTER-E - Energy Consumption Data for the BUTTER Empirical Deep Learning Dataset".  United States.  doi:https://doi.org/10.25984/2329316.  https://www.osti.gov/servlets/purl/2329316. Pub date:Fri Dec 30 04:00:00 UTC 2022

Copy to clipboard


                    
@article{osti_2329316,

  title        = {BUTTER-E - Energy Consumption Data for the BUTTER Empirical Deep Learning Dataset},

  author       = {Tripp, Charles and Perr-Sauer, Jordan and Bensen, Erik and Gafur, Jamil and Nag, Ambarish and Purkayastha, Avi},

  abstractNote = {The BUTTER-E - Energy Consumption Data for the BUTTER Empirical Deep Learning Dataset adds node-level energy consumption data from watt-meters to the primary sweep of the BUTTER - Empirical Deep Learning Dataset. This dataset contains energy consumption and performance data from 63,527 individual experimental runs spanning 30,582 distinct configurations: 13 datasets, 20 sizes (number of trainable parameters), 8 network "shapes", and 14 depths on both CPU and GPU hardware collected using node-level watt-meters. This dataset reveals the complex relationship between dataset size, network structure, and energy use, and highlights the impact of cache effects. BUTTER-E is intended to be joined with the BUTTER dataset (see "BUTTER - Empirical Deep Learning Dataset on OEDI" resource below) which characterizes the performance of 483k distinct fully connected neural networks but does not include energy measurements.},

  doi          = {10.25984/2329316},

  journal      = {},

  number       = ,

  volume       = ,

  place        = {United States},

  year         = {Fri Dec 30 04:00:00 UTC 2022},

  month        = {Fri Dec 30 04:00:00 UTC 2022}

}

Copy to clipboard

Dataset:

View Dataset

DOI: https://doi.org/10.25984/2329316

Save / Share:

Export Metadata

Save to My Library

Similar records in DOE Data Explorer and OSTI.GOV collections:

BUTTER - Empirical Deep Learning Dataset

Dataset Tripp, Charles ; Perr-Sauer, Jordan ; Hayne, Lucas ; ...

The BUTTER Empirical Deep Learning Dataset represents an empirical study of the deep learning phenomena on dense fully connected networks, scanning across thirteen datasets, eight network shapes, fourteen depths, twenty-three network sizes (number of trainable parameters), four learning rates, six minibatch sizes, four levels of label noise, and fourteen levels of L1 and L2 regularization each. Multiple repetitions (typically 30, sometimes 10) of each combination of hyperparameters were preformed, and statistics including training and test loss (using a 80% / 20% shuffled train-test split) are recorded at the end of each training epoch. In total, this dataset covers 178 thousandmore »« less
ARPA-E PERFORM datasets

Dataset Sergi, Brian ; Feng, Cong ; Zhang, Flora ; ...

Time-coincident load, wind, and solar data including actual and probabilistic forecast datasets at 5-min resolution for ERCOT, MISO, NYISO, and SPP. Wind and solar profiles are supplied for existing sites as well as planned sites based on interconnection queue projects as of 2021. For ERCOT actuals are provided for 2017 and 2018 and forecasts for 2018, and for the remaining ISOs actuals are provided for 2018 and 2019 and forecasts for 2019. There datasets were produced by NREL as part of the ARPA-E PERFORM project, an ARPA-E funded program that aim to use time-coincident power and load seeks to developmore »« less
Dataset, Code, and Models for Training Deep Learning Potentials for Low Temperature Plasma-Surface Interactions

Dataset Draney, Jack S. ; Panagiotopoulos, Athanassios ; Graves, David

This repository contains datasets, training scripts, and finished models, and test simulations used in the development of DeepREBO— a machine-learned interatomic potential trained to emulate the REBO2 empirical potential. The data was generated to study deep potential development for simulations of plasma-surface interactions. It uses an active learning framework, starting from a minimal dataset and iteratively expanding it. Included are those generated datasets, the trained models, and simulations used to evaluate the performance of the training process. This resource supports reproducibility and provides a reference framework for training deep potentials in plasma-surface interaction studies.
Data for The utility of transfer learning to improve the performance of deep learning in axon segmentation

Dataset Oostrom, Marjolein T ; Muniak, Michael ; Eichler West, Rogene M ; ...

The utility of transfer learning to improve the performance of deep learning in axon segmentation Data Data: All the input and labeled volumes tf-logs: Tensorflow logs, view with command "tensorboard --logdir [name of folder]" Model Weights: model_weights: the argument list under variable combo indicate 1) no oversampling, 2) no rotation, 3) no learn scheduler, and 4) flipping on all three dimensions, and the additional values indicate 5) elastic deformation percentage, 6) rotate deformation percentage, 7) layer setting , 8) learning rate, and 9) training/validation/test data division suffix (leave '' if not using suffix). Results: Output from inference segment_total_results_validation_final: All validationmore »« less
OPFLearnData: Dataset for Learning AC Optimal Power Flow

Dataset Joswig-Jones, Trager ; Zamzam, Ahmed ; Baker, Kyri

The datasets are resulting from OPFLearn.jl, a Julia package for creating AC OPF datasets. The package was developed to provide researchers with a standardized way to efficiently create AC OPF datasets that are representative of more of the AC OPF feasible load space compared to typical dataset creation methods. The OPFLearn dataset creation method uses a relaxed AC OPF formulation to reduce the volume of the unclassified input space throughout the dataset creation process. The dataset contains load profiles and their respective optimal primal and dual solutions. Load samples are processed using AC OPF formulations from PowerModels.jl. More information onmore »« less

Similar Records