MLflow Server

Self-hosted MLflow tracking server with authentication, designed for Coolify deployment.

Features

• SQLite backend for experiment tracking
• Artifact storage with proxy serving
• Basic authentication with auto-generated admin password
• Multi-user support with permissions (READ, EDIT, MANAGE)
• Persistent named volume for data

Quick Start (Local)

docker compose up -d
# Access at http://localhost:5001
# Default credentials: admin / password1234

Coolify Deployment

1. Create a new service in Coolify using this repo
2. Set domain in Coolify UI (e.g., https://yourdomain.com/mlflow:5000)
3. Coolify auto-generates SERVICE_PASSWORD_MLFLOW for the admin password
4. Find the password in Coolify's environment variables dashboard

User Management

Creating Users

Option 1: Signup page Visit https://your-mlflow-domain/signup to create a new account.

Option 2: Admin API

curl -X POST -u admin:$ADMIN_PASSWORD \
  "https://your-mlflow-domain/api/2.0/mlflow/users/create" \
  -H "Content-Type: application/json" \
  -d '{"username": "teammate", "password": "their-password"}'

Option 3: Python

from mlflow.server import get_app_client

auth_client = get_app_client("basic-auth", tracking_uri="https://your-mlflow-domain/")
auth_client.create_user(username="teammate", password="their-password")

Permissions

Permission	Can Read	Can Update	Can Delete	Can Manage
READ	Yes	No	No	No
EDIT	Yes	Yes	No	No
MANAGE	Yes	Yes	Yes	Yes

• Default permission: READ (all users can view experiments)
• Experiment creators: Automatically get MANAGE permission
• Admin users: Unrestricted access to everything

Granting Permissions

from mlflow.server import get_app_client

auth_client = get_app_client("basic-auth", tracking_uri="https://your-mlflow-domain/")
auth_client.create_experiment_permission(
    experiment_id="1",
    username="teammate",
    permission="EDIT"  # or "MANAGE"
)

Promoting to Admin

auth_client.update_user_admin(username="teammate", is_admin=True)

Client Configuration

export MLFLOW_TRACKING_URI=https://your-mlflow-domain
export MLFLOW_TRACKING_USERNAME=your-username
export MLFLOW_TRACKING_PASSWORD=your-password

Or use ~/.mlflow/credentials:

[mlflow]
mlflow_tracking_username = your-username
mlflow_tracking_password = your-password

Data Persistence

All data is stored in the mlflow-data named volume:

• mlflow.db - experiments, runs, metrics, parameters
• artifacts/ - uploaded files
• auth.db - user accounts and permissions