Deploy MLflow
Track ML experiments, models and runs with a web UI
MLflow
Just deployed
/data
Deploy and Host MLflow on Railway
MLflow is an open-source platform for tracking machine-learning experiments, packaging models and managing their lifecycle.
About Hosting MLflow
This template runs the official ghcr.io/mlflow/mlflow image as a tracking server with basic authentication enabled. Metadata is kept in SQLite and artifacts in a persistent volume at /data. The admin user is admin; its password is the generated MLFLOW_ADMIN_PASSWORD variable on the service.
Common Use Cases
- Log parameters, metrics and artifacts from training runs
- Compare experiments in the web UI
- Register and version models
Dependencies for MLflow Hosting
- One service with a volume; nothing else
Deployment Dependencies
Why Deploy MLflow on Railway?
Railway gives you a public HTTPS URL, a persistent volume and automatic restarts with no server to maintain. Point your code at the public URL and set MLFLOW_TRACKING_USERNAME=admin and MLFLOW_TRACKING_PASSWORD to the generated password.
Template Content
