AWS-AutoML-Lite/docker-compose.yml at main · cristofima/AWS-AutoML-Lite · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
# Docker Compose for local development
# Uses real AWS services (dev environment) for testing before deployment
#
# Prerequisites:
# 1. AWS credentials configured (~/.aws/credentials or environment variables)
# 2. Infrastructure deployed to dev environment (terraform apply)
# 3. Copy .env.example to .env and fill in values from terraform output
#
# Usage:
#   docker-compose up                               # Start API only
#   docker-compose --profile training run training  # Run training job
#
# Frontend: Run natively with pnpm (see frontend/README.md)
#   cd frontend && pnpm dev
#
# For training, set environment variables inline:
#   DATASET_ID=xxx TARGET_COLUMN=price docker-compose --profile training run training

services:
  # Backend API - connects to AWS dev environment
  api:
    build:
      context: ./backend
      dockerfile: Dockerfile.api
    container_name: automl-api
    ports:
      - "8000:8000"
    env_file:
      - ./backend/.env
    environment:
      - AWS_PROFILE=${AWS_PROFILE:-default}
      # Map AWS_REGION to REGION for training container compatibility
      - REGION=${AWS_REGION:-us-east-1}
    volumes:
      - ./backend/api:/app/api:ro
      - ~/.aws:/root/.aws:ro  # Mount AWS credentials
    # Required for ONNX Runtime in Docker (development only)
    # Lambda doesn't need this - works natively
    privileged: true
    command: uvicorn api.main:app --host 0.0.0.0 --port 8000 --reload

  # Training Container - same image used by AWS Batch
  # Run manually: docker-compose --profile training run training
  training:
    build:
      context: ./backend/training
      dockerfile: Dockerfile
    container_name: automl-training
    profiles:
      - training  # Only runs when explicitly called with --profile training
    env_file:
      - ./backend/.env
    environment:
      - AWS_PROFILE=${AWS_PROFILE:-default}
      # Map AWS_REGION to REGION for train.py compatibility
      - REGION=${AWS_REGION:-us-east-1}
      # These must be set when running (inline):
      - DATASET_ID=${DATASET_ID:-}
      - TARGET_COLUMN=${TARGET_COLUMN:-}
      - JOB_ID=${JOB_ID:-local-test}
      - TIME_BUDGET=${TIME_BUDGET:-60}
    volumes:
      - ~/.aws:/root/.aws:ro  # Mount AWS credentials