Skip to content

Latest commit

 

History

History
45 lines (34 loc) · 865 Bytes

running-bigdl.md

File metadata and controls

45 lines (34 loc) · 865 Bytes

Sample Lab : Running BigDL

About BigDL

BigDL is a deep learning framework that runs on Spark

Screencast

Coming soon

Checkout our Sandbox channel for more videos.

Running BigDL

Environment variables

BIGDL_HOME=/usr/local/apps/bigdl
SPARK_HOME=$HOME/spark

Step 1 : Install Spark

In terminal

  $   cd
  $   tar xvf files/spark-2.2.0-bin-hadoop2.7.tgz
  $   mv  spark-2.2.0-bin-hadoop2.7  spark

Get the tutorials

  $  git clone https://github.com/elephantscale/bigdl-tutorials

Run BigDL

    $  cd  bigdl-tutorials
    $  ./run-bigdl-native.sh
  • This will launch Jupyter notebook.
  • Note the URL printed on console
  • Go to the URL in browser
  • sample notebooks are in notebooks directory
  • have fun!