Public Repository

Last pushed: a year ago
Short Description
Parallelizing Stochastic Gradient Descent for Deep Convolutional Neural Network
Full Description

We provide here a docker image for the implementation of our paper "SpeeDO: Parallelizing Stochastic Gradient Descent for Deep Convolutional Neural Network" , Zheng, Zhongyang and Jiang, Wenrui and Wu, Gang and Chang, Edward Y.

As a quick start to run system under docker env ( master and 3 workers ), you can

  1. Pull docker image on master and worker instances

    docker pull openbigdatagroup/speedo:latest
    
  2. Start master process on master instance

    docker run -d --name=speedo-master --net=host obdg/speedo master <master-address> 3 --test 0 --maxIter 1000 --movingRate 0.5
    
  3. Start worker process on worker instances

    docker run -d --name=speedo-worker --net=host obdg/speedo worker <master-address> <worker-address>
    

For detail info, please refer to http://openbigdatagroup.github.io/speedo/

Docker Pull Command
Owner
obdg

Comments (0)