Training MLOps 1

This is a practical work to illustrate Data Science industrialization. It is part of a training provided by Octo Academy, more details here

How do I install it

First, clone the project from GitLab : (if you are on Windows you will need Git for windows and do git command in Git Bash)

git clone git@github.com:octo-technology/Formation-MLOps-1.git

Second, open project in Pycharm.

If on Windows, configure your terminal in Pycharm so that you can run all commands :

Go to Settings > Tools > Terminal
Change “Shell path” by : cmd.exe "/K" "C:\Users\>>me<<\Miniconda3\Scripts\activate.bat"
Restart Pycharm
Test it by typing git in terminal

Third, make sure you have miniconda or anaconda installed. If not, install it!

Change directory into the repo you cloned

cd Formation-MLOps-1

Create a conda env

conda create -n formation_mlops_1 python=3.10

Activate your env

conda activate formation_mlops_1

Install all needed dependencies

pip install -r requirements.txt

Start a jupyter notebook in the folder

jupyter-notebook

If your formation_mlops_1 environment is not available in jupyter interface (when clicking on new). You should :

It is highly linked to the presentation of the formation.

To navigate between steps change branch.

To see all branches

git branch -a

To start the practical work you should check out branch 0_initial_state

git checkout 0_initial_state