OpenMC-LOBOS/docker-compose.yml at main · light0vij/OpenMC-LOBOS · 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
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
# =============================================================================
# docker-compose.yml
#
# ACCESS:
#   http://localhost:8888    — no password needed
#
# LLM API KEY:
#   The notebook reads its LLM key from the LLM_API_KEY environment variable
#   (never hardcode a real key in the notebook). Set it before "docker
#   compose up", e.g.:
#     LLM_API_KEY=your_key_here docker compose up
#   or put LLM_API_KEY=your_key_here in a local .env file next to this file
#   (docker compose loads .env automatically; .env is already git-ignored).
# =============================================================================

services:

  openmc-jupyter:

    image: ghcr.io/light0vij/openmc-lobos:latest

    container_name: openmc-lobos

    ports:
      - "8888:8888"

    volumes:

      # Mount Notebooks/ so notebooks (and any you add) persist on the host
      # and are browsable/editable from JupyterLab's file tree.
      - ./Notebooks:/workspace/Notebooks

      # Mount environments/ so the model-specific OpenMC wrappers imported by
      # the notebooks are visible and editable from JupyterLab too.
      - ./environments:/workspace/environments

      # Mount openmc_bo/ (LLM sampler, GP/EI, BO loop, LLM bounds
      # orchestrator) so it's browsable/editable from JupyterLab too.
      - ./openmc_bo:/workspace/openmc_bo

      # Mount sampler_prompts/ so the LLM Generative Seeding prompt configs
      # can be tweaked without rebuilding the image.
      - ./sampler_prompts:/workspace/sampler_prompts

      # Mount prompt2.json — the LLM Bounds Orchestration prompt config —
      # for the same reason.
      - ./prompt2.json:/workspace/prompt2.json

      # OpenMC result files (statepoints, tallies, plots, CSVs, reports)
      - ./Results:/workspace/Results

      # Nuclear data — stored in a named Docker volume so it only downloads
      # once and persists across container restarts.
      # To use existing data instead, replace with:
      # - /path/to/your/nuclear_data:/nuclear_data
      - nuclear_data:/nuclear_data

    environment:
      - OPENMC_CROSS_SECTIONS=/nuclear_data/cross_sections.xml
      - JUPYTER_ENABLE_LAB=yes
      # Passed straight through to the notebook's os.environ.get("LLM_API_KEY").
      - LLM_API_KEY=${LLM_API_KEY:-}

    # Adjust cpus to how many cores you want to dedicate to OpenMC runs.
    # Remove deploy block entirely to use all available resources.
    deploy:
      resources:
        limits:
          cpus: "4"
          memory: "8g"

    restart: unless-stopped

    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:8888/lab"]
      interval: 30s
      timeout: 10s
      retries: 3
      start_period: 30s

volumes:
  nuclear_data:
    name: neutronics_nuclear_data