Skip to main content
Welcome to this comprehensive guide on setting up an MLflow server for your local development environment using Visual Studio Code. Follow along step-by-step with your terminal while you watch the terminal pane in VS Code similar to the demonstration below.

​
Step 1: Install MLflow

MLflow is available as a package on PyPI. To install MLflow, run the following command in your terminal:
Upon executing this command, you should observe output similar to the example below:
Additionally, you may encounter more package download messages, like in the output below:
If the installation process prompts any errors, ensure that your pip version is up-to-date and you are using a supported version of Python.

​
Step 2: Start the MLflow Server

Once the installation is complete, launch the MLflow UI by running:
This command will start the MLflow server, which listens on the default port 5000. You should see terminal output that resembles this:
Click on the provided link or manually navigate to http://127.0.0.1:5000 in your browser to access the MLflow UI.

​
Step 3: Explore the MLflow UI

When you open the MLflow UI in your browser, you will initially see a default experiment. If no experiments have been logged, the experiment section will appear empty.
The image shows an MLflow interface with no logged runs in the "Experiments" section, displaying a message that no runs have been logged yet.
The MLflow UI acts as a centralized hub where all experiments are recorded. Here, you can:
  • Begin a new experiment with a custom name.
  • View detailed logs and traces.
  • Compare the performance of different runs.
  • Access the dedicated models section, which serves as the MLflow model repository.
In this demonstration, MLflow version 2.17.2 was installed by default since no specific version was provided during installation.
The image shows an MLflow interface with a "Registered Models" page, indicating no models are registered yet and offering an option to create a model.

​
Conclusion

This tutorial has demonstrated how to set up a basic MLflow service on a local system. This setup can also be applied to other environments such as virtual machines or Kubernetes clusters. Thank you for following this guide. Stay tuned for more detailed tutorials on advanced MLflow functionalities and usage scenarios. Happy experimenting!

Watch Video