recommendation-system/docker-compose.yaml at main · amit-chaubey/recommendation-system · 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
# MLOps-style deploy: API (model serving) + UI (Streamlit).
# Build from repo root:
#   docker compose build
# Run:
#   docker compose up
# Optional: run pipeline to (re)train before up:
#   docker compose run --rm api python run_pipeline.py --config configs/model_config.yaml
services:
  api:
    build:
      context: .
      dockerfile: Dockerfile
    ports:
      - "8000:8000"
    environment:
      - TMDB_API_KEY=${TMDB_API_KEY:-}
    volumes:
      # Mount trained models so you can update without rebuild (e.g. after run_pipeline)
      - ./models/trained:/app/models/trained:ro

  ui:
    build:
      context: .
      dockerfile: ui/Dockerfile
    ports:
      - "8501:8501"
    environment:
      - TMDB_API_KEY=${TMDB_API_KEY:-}
    volumes:
      - ./models/trained:/app/models/trained:ro
      - ./.env.dev:/app/.env.dev:ro
    depends_on:
      - api