An example on how to train a Convolutional neural network
on Apache Spark.
Feel free to fork and contribute!
This repository aims to be an easy and well documented example on how to train a Convolutional neural network
taking advantage of Apache Spark's distributed computation.
It uses Deeplearning4j on top of Apache Spark.
The neural network is trained against digit classifications from the MNIST data set,
a subset of a larger set available from NIST (National Institute of Standards and Technology).
This example is set up to run on a local instance of Spark (i.e., it's runnable within your IDE).
To submit and run it on a cluster, remove the .setMaster() method and use Spark submit.
This example is a fork of dl4j-spark-cdh5-examples.
A joint effort between ebezzi (Emanuele Bezzi) and Vincibean (Andrea Bessi)
Unless stated elsewhere, all files herein are licensed under the GPLv3 license. For more information, please see the LICENSE file.