guesemha

A multiprocess MATLAB toolbox for parallel hyperparamater tuning for Machine Learning applications
6 Downloads
Aktualisiert 17. Apr 2022

guesemha View guesemha on File Exchange

A multiprocess MATLAB toolbox for parallel hyperparamater tuning for Machine Learning applications

Installation

Requirements

  • SharedMatrix Allows any Matlab object to be shared between Matlab sessions (w/o using file I/O). Used to share data accross Matlab workers.
  • Instrument Control Toolbox, used for IPC.

A compiled 64-bit mex function of SharedMatrix is shipped with the toolbox.

-Download or clone the toolbox
-cd to guesemha folder
-in the MATLAB command window run:
>> setup

Usage

see demo scripts in examples folder.

Support

test on Windows 7 64bit on MATLAB releases:

  • R2014a
  • R2015a

for support, open an issue detailing the problem you encountered with the toolbox.

Limitations

Works only on multicore systems with local memory

Contributing

Pull requests are welcome, open an issue first to discuss proposed changes.

Licence

MIT

Acknowledgement

I would like to thank the SharedMatrix developer Joshua Dillon and Andrew Smith for the windows version.

Zitieren als

Okba Bekhelifi (2024). guesemha (https://github.com/okbalefthanded/guesemha/releases/tag/0.0.1.0), GitHub. Abgerufen .

Kompatibilität der MATLAB-Version
Erstellt mit R2014a
Kompatibel mit allen Versionen
Plattform-Kompatibilität
Windows macOS Linux
Tags Tags hinzufügen
Quellenangaben

Inspiriert von: sharedmatrix

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

dependencies/CStrCatStr

dependencies/fRMField

dependencies/jsystem/matlab-jsystem/src

dependencies/jsystem/matlab-jsystem/test

dependencies/sharedmatrix

dependencies/sharedmatrix/sharedmatrix_windows_by_andrew_smith

examples

Version Veröffentlicht Versionshinweise
0.0.1.0

Um Probleme in diesem GitHub Add-On anzuzeigen oder zu melden, besuchen Sie das GitHub Repository.
Um Probleme in diesem GitHub Add-On anzuzeigen oder zu melden, besuchen Sie das GitHub Repository.