- README.md: Project Description
- main.ipynb: Jupyter Notebook with MLflow code example
This is the companion respository for the Heartbeat article Enhancing MLflow with Comet by Matt Blasa.
The contents of this repository is example source code demonstrating Comet's machine learning and experiment tracking used for MLflow machine learning experiments.
Click below for the Medium article:
https://medium.com/fritzheartbeat/enhancing-mlflow-with-comet-a39def1954e8
The data used in this project is taken from the University of Californina Irvine wine quality dataset which can be found here: https://archive.ics.uci.edu/ml/datasets/wine
This project is focused for the Azure Databricks environment. To run this project, you will need a Databricks to install Comet library to your cluster, and upload the project file.
Using Project File
NOTE: that Comet API Key has been removed from project file, for privacy reasons. You will need to add your experiment API key to use part of this code
- Comet Experiment: https://www.comet.ml/mattblasa/enhancing-ml-flow/view/new/panels
- Comet Documentation: https://www.comet.ml/docs/python-sdk/Experiment/