version: '3.8'

services:
  backend:
    build:
      context: ./backend
      args:
        - --no-cache
    ports:
      - "8000:80"
    networks:
      - app-network
    volumes:
      - .env:/app/.env
      - ./data:/app/data
      - ./workflow_artifacts:/app/workflow_artifacts
      - ~/.azure:/root/.azure
      - ./mlruns:/app/mlruns:rw
      - ./mlartifacts:/app/mlartifacts:rw

  frontend:
    build:
      context: ./frontend
      args:
        - --no-cache
    ports:
      - "8501:8501"
    networks:
      - app-network

  mlflow:
    image: ghcr.io/mlflow/mlflow:v2.16.0
    ports:
      - "5000:5000"
      - "8080:8080"
    networks:
      - app-network
    volumes:
      - ./mlruns:/mlruns:rw
      - ./mlartifacts:/mlartifacts:rw
    environment:
      MLFLOW_LOGGING_LEVEL: "DEBUG"
#      MLFLOW_TRACKING_URI: "http://mlflow:8080"
#      MLFLOW_BACKEND_STORE_URI: "sqlite:///mlflow/mlruns/mlruns.db"
#      MLFLOW_DEFAULT_ARTIFACT_ROOT: "/mlflow/mlruns"
    command: >
      mlflow server --host 0.0.0.0 --port 8080
      --backend-store-uri sqlite:///mlruns/mlruns.db
#      --default-artifact-root file:///mlruns

networks:
  app-network: