Beginning MLOps

Data

Get the data from Kaggle website here and save it inside the data folder

Prepare

Install the poetry environment from the project base folder

poetry install --no-root

Run

This time we have two folders that have very similar scripts but with one particular difference. Run either one of them (finishing the other if started)

beginning mlops

Here from one terminal run

cd beginningmlops
poetry run mlflow ui

and then from another terminal run

cd beginningmlops
poetry run python practice1MLFlow.py

With that when you go to http://localhost:5000/ you will see the MLOps UI and can see the runs on the experiment scikit_learn_experiment

servermlops

Here run from one terminal

cd servermlops
poetry run mlflow server --backend-store-uri sqlite:///mlflow.db --default-artifact-root ./mlruns --host 0.0.0.0 --port 5000

and then from another terminal run

cd servermlops
poetry run python practice1MLFlow.py

Here to by going to http://localhost:5000/ you will see the MLOps UI and can see the runs on the experiment scikit_learn_experiment

In order to learn the difference go to the Medium article here