Set Up MLflow Server

MLflow is open source, and you can set up the MLflow server using either pip or docker.

Before you can leverage MLflow for your GenAI application development, you must first start the MLflow server.

Local (uv)

Install the Python package manager uv (that will also install uvx command to invoke Python tools without installing them).

Start a MLflow server locally.

shell

uvx mlflow server

Local (pip)

Python Environment: Python 3.10+

Install the mlflow Python package via pip and start a MLflow server locally.

shell

pip install --upgrade mlflow
mlflow server

Local (docker)

MLflow provides a Docker Compose file to start a local MLflow server with a PostgreSQL database and a MinIO server.

shell

git clone --depth 1 --filter=blob:none --sparse https://github.com/mlflow/mlflow.git
cd mlflow
git sparse-checkout set docker-compose
cd docker-compose
cp .env.dev.example .env
docker compose up -d

Refer to the instruction for more details (e.g., overriding the default environment variables).

This will start the server at port 5000 on your local machine and you can access the MLflow web UI at http://localhost:5000.

If you are looking for more guidance about self-hosting the MLflow server, please see the Self-Hosting Guide for more details.

info

If you are using MLflow on Databricks, please visit this for environment setup instructions specific to Databricks.

Next Step

Now that you have started the MLflow server, let's start tracing your GenAI application.

Follow this quickstart to send your GenAI application traces to the MLflow server.