JustInCase/docker-compose.yml at main · companionintelligence/JustInCase · 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
services:
  # ── Shared build (both server and ingestion use the same image) ────
  jic:
    build:
      context: .
      dockerfile: Dockerfile
    image: jic:latest
    container_name: jic-server
    deploy:
      resources:
        limits:
          cpus: '0'       # no CPU limit — use all available cores
          memory: 8G
    volumes:
      - ./public:/app/public
      - ./data:/app/data
      - ./gguf_models:/app/gguf_models:ro
    ports:
      - "8080:8080"
    environment:
      - LLM_MODEL=llama3.2:3b
      - EMBEDDING_MODEL=nomic-embed-text
      - LLM_GGUF_FILE=Llama-3.2-3B-Instruct-Q4_K_M.gguf
      - EMBEDDING_GGUF_FILE=nomic-embed-text-v1.5.Q4_K_M.gguf
    command: ["./jic-server"]
    healthcheck:
      test: ["CMD-SHELL", "echo > /dev/tcp/localhost/8080 || exit 1"]
      interval: 30s
      timeout: 5s
      start_period: 60s
      retries: 3

  ingestion:
    image: jic:latest          # reuses the image built by the jic service
    container_name: jic-ingestion
    depends_on:
      jic:
        condition: service_started
    deploy:
      resources:
        limits:
          cpus: '2'
          memory: 4G
    volumes:
      - ./public:/app/public
      - ./data:/app/data
      - ./gguf_models:/app/gguf_models:ro
    environment:
      - LLM_MODEL=llama3.2:3b
      - EMBEDDING_MODEL=nomic-embed-text
      - LLM_GGUF_FILE=Llama-3.2-3B-Instruct-Q4_K_M.gguf
      - EMBEDDING_GGUF_FILE=nomic-embed-text-v1.5.Q4_K_M.gguf
    command: ["./jic-ingestion"]