Example of usage of MLFlow tracking

Repo contains two notebooks which demonstrate functionality of MLFlow tracking:

/notebooks/exp_mlflow_local_autolog.ipynb: basic use with autlog() from MLflow

/notebooks/exp_mlflow_serve_custom.ipynb: custom configuration of tracking parameters

More detailed information about model can be found in this repo and blog post

Step 1: create conda environment

conda create -n mlflow_tracking-env python=3.9

conda activate mlflow_tracking-env

Step 2: install requirements

pip install -r requirements.txt

Step 3: launch MLFlow UI

Option 1: without using tracking server

cd /notebooks #mlflow should be always launched from the folder with notebooks/scripts

mlflow ui

Go to http://127.0.0.1:5000 which will open UI for tracked experiments.

Option 2: with local tracking server

cd /notebooks #mlflow should be always launched from the folder with notebooks/scripts

mlflow server --backend-store-uri sqlite:///backend.db --default-artifact-root ./artifacts_local

Step 4: run notebook

Sources for MLflow setups:

MLOps Zoomcamp 2.2

MLOps Zoomcamp 2.6