Setting up an Azure Ubuntu VM for Deep Learning

Step1.sh

wget https://raw.githubusercontent.com/fastai/courses/master/setup/install-gpu-azure.sh
bash install-gpu-azure.sh

step2.sh

source .bashrc
jupyter notebook --generate-config
jupass=`python -c "from notebook.auth import passwd; print(passwd())"`
echo "c.NotebookApp.password = u'"$jupass"'" >> .jupyter/jupyter_notebook_config.py
echo "c.NotebookApp.ip = '*'
c.NotebookApp.open_browser = False" >> .jupyter/jupyter_notebook_config.py
mkdir nbs

step3_from_your_local

ssh -L 8000:localhost:8888 ubuntu@13.92.246.242 

step4.sh

cd nbs
wget https://raw.githubusercontent.com/fastai/courses/master/deeplearning1/nbs/lesson1.ipynb
wget http://www.platform.ai/files/nbs/utils.zip
wget http://www.platform.ai/files/nbs/vgg16.zip
sudo apt install unzip
unzip -q utils
unzip -q vgg16
#At this moment you should be in ~/nbs folder
mkdir data
cd data
wget http://www.platform.ai/files/dogscats.zip
unzip -q dogscats.zip
cd..
jupyter notebook

step5.sh

#Source : http://stackoverflow.com/questions/25215102/installing-opencv-for-python-on-ubuntu-getting-importerror-no-module-named-cv2
conda install opencv

thanks for your article - it is a great help.
fyi www.platform.ai now needs to be replaced with files.fast.ai.
I'd PR it but can't since it is a gist

Hey Abhik - have you figured out if it is possible to directly connect to the ubuntu machine without SSH tunneling?

@isaacy : Edit your Network Security Group to enable TCP inbound on port 8888.

Ref: http://wiki.fast.ai/index.php/Azure_install