Deep Learning on Spark tutorial
docker build -t bythebay/spark-deep-learning:0.0.1 .This uses a variable LOCAL_NOTEBOOKS which refers to a local directory containing the notebooks you want to include and keep up to date during the session.
Another folder you might want to sync is the data dir, which uses LOCAL_DATA then.
export LOCAL_NOTEBOOKS=<path to local notebooks dir>
export LOCAL_DATA=<path to local data dir>
docker run -v $LOCAL_NOTEBOOKS:/root/spark-notebook/notebooks/spark-deep-learning \
-v $LOCAL_DATA:/root/data/spark-deep-learning \
--rm -it -m 8g \
-p 19000:9000 \
-p 14040:4040 -p 14041:14041 -p 14042:4042 -p 14043:14043 \
bythebay/spark-deep-learning:0.0.1 \
bashThen
source var.sh
source start.sh
source create.shOpen browser at http://localhost:19000/tree/spark-deep-learning.
WARN:
If you're running docker via a VM (Mac, Msft, ...) then you need to replace
localhostby the VM's IP.