usage.md

Usage

Once you have ML Studio installed, find out its the IP address to access it, login and start building.

Find the IP to access ML Studio

kubectl get svc istio-ingressgateway -n istio-system

Log in as an admin user

  USER: admin
  PASSWORD: password

Work directory

{% hint style="info" %} ML Studio root work directory can be access at /home/jovyan/work inside jupyter pods and /home/work inside other pods.

All files you would like persisted should live under this directory. {% endhint %}

How to manage your python environment

We recommend using conda to manage all your conda environment but you are free to use others.

{% hint style="warning" %} base conda environment will not get persisted {% endhint %}

Create a new environment

Start a new terminal session under Notebooks

conda create --name py3 python=3.7 -y

Add it as a new Jupyter kernel

Before you can use this new python enviorment in your notebooks, you need to make Jupyter aware of it first.

conda activate test                   

ipython kernel install --user --name=py3

Finally, refresh ML Studio web app for your changes to take effect.