Integration

Python Client Integration

python

# mlflow_client.py
import mlflow
import mlflow.sklearn
from mlflow.tracking import MlflowClient
# Configure MLflow
mlflow.set_tracking_uri("http://mlflow.example.com")
mlflow.set_experiment("production-models")
# Initialize client
client = MlflowClient()
# Example: Log experiment
with mlflow.start_run():
    # Log parameters
    mlflow.log_param("learning_rate", 0.01)
    mlflow.log_param("max_depth", 5)
    # Log metrics
    mlflow.log_metric("accuracy", 0.95)
    mlflow.log_metric("precision", 0.92)
    # Log model
    mlflow.sklearn.log_model(
        sk_model=model,
        artifact_path="model",
        registered_model_name="production-classifier"
    )

REST API Integration

bash

# Create experiment
curl -X POST "http://mlflow.example.com/api/2.0/mlflow/experiments/create" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "api-experiment",
    "artifact_location": "/opt/mlflow/artifacts/api-experiment"
  }'
# Start run
curl -X POST "http://mlflow.example.com/api/2.0/mlflow/runs/create" \
  -H "Content-Type: application/json" \
  -d '{
    "experiment_id": "1",
    "start_time": 1640995200000,
    "tags": [{"key": "environment", "value": "production"}]
  }'
# Log metrics
curl -X POST "http://mlflow.example.com/api/2.0/mlflow/runs/log-metric" \
  -H "Content-Type: application/json" \
  -d '{
    "run_id": "run-123",
    "key": "accuracy",
    "value": 0.95,
    "timestamp": 1640995200000
  }'

CI/CD Integration

yaml

# .github/workflows/mlflow-deploy.yml
name: MLflow Model Deployment
on:
  push:
    branches: [main]
jobs:
  deploy:
    runs-on: ubuntu-latest
    steps:
    - uses: actions/checkout@v3
    - name: Set up Python
      uses: actions/setup-python@v4
      with:
        python-version: '3.10'
    - name: Install dependencies
      run: |
        pip install mlflow==3.0.1
        pip install -r requirements.txt
    - name: Train and log model
      env:
        MLFLOW_TRACKING_URI: ${{ secrets.MLFLOW_TRACKING_URI }}
        MLFLOW_TRACKING_USERNAME: ${{ secrets.MLFLOW_USERNAME }}
        MLFLOW_TRACKING_PASSWORD: ${{ secrets.MLFLOW_PASSWORD }}
      run: |
        python train_model.py
        python deploy_model.py

Kubernetes Integration

yaml

# mlflow-deployment.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
  name: mlflow-server
spec:
  replicas: 3
  selector:
    matchLabels:
      app: mlflow-server
  template:
    metadata:
      labels:
        app: mlflow-server
    spec:
      containers:
      - name: mlflow
        image: mlflow:3.0.1
        ports:
        - containerPort: 5000
        env:
        - name: MLFLOW_BACKEND_STORE_URI
          valueFrom:
            secretKeyRef:
              name: mlflow-secrets
              key: database-uri
        - name: MLFLOW_DEFAULT_ARTIFACT_ROOT
          value: "s3://mlflow-artifacts"
        resources:
          requests:
            memory: "1Gi"
            cpu: "500m"
          limits:
            memory: "2Gi"
            cpu: "1000m"
Related content