version: '3.8'

services:
  # KerdosAI Training Service
  kerdosai-train:
    build:
      context: .
      target: development
    image: kerdosai:dev
    container_name: kerdosai-train
    volumes:
      - .:/app
      - ./data:/app/data
      - ./output:/app/output
      - ./checkpoints:/app/checkpoints
    environment:
      - PYTHONPATH=/app
      - WANDB_API_KEY=${WANDB_API_KEY}
    command: python cli.py train --config configs/default.yaml
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: all
              capabilities: [gpu]

  # KerdosAI API Service
  kerdosai-api:
    build:
      context: .
      target: production
    image: kerdosai:prod
    container_name: kerdosai-api
    ports:
      - "8000:8000"
    volumes:
      - ./output:/app/output:ro
    environment:
      - PYTHONPATH=/app
      - MODEL_PATH=/app/output
    restart: unless-stopped
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:8000/health"]
      interval: 30s
      timeout: 10s
      retries: 3
      start_period: 40s

  # TensorBoard for monitoring
  tensorboard:
    image: tensorflow/tensorflow:latest
    container_name: kerdosai-tensorboard
    ports:
      - "6006:6006"
    volumes:
      - ./runs:/logs:ro
    command: tensorboard --logdir=/logs --host=0.0.0.0
    restart: unless-stopped

networks:
  default:
    name: kerdosai-network