FinetuneOrch/docker-compose.yml at main · SunPCSolutions/FinetuneOrch · 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
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
services:
  backend:
    build: ./backend
    ports:
      - "7000:8000"
    volumes:
      - ./backend/src:/app
      - ./saves:/app/saves
      - uploads:/app/uploads
      - /var/run/docker.sock:/var/run/docker.sock
    environment:
      - OLLAMA_API_URL=${OLLAMA_API_URL}
      - LLAMA_FACTORY_CONTAINER=${LLAMA_FACTORY_CONTAINER}
      - LLAMA_CPP_CONTAINER=${LLAMA_CPP_CONTAINER}
      - OLLAMA_CONTAINER=${OLLAMA_CONTAINER}
    networks:
      - docknet

  frontend:
    build:
      context: ./frontend
      dockerfile: Dockerfile
      args:
        - VITE_LLAMA_FACTORY_URL=${VITE_LLAMA_FACTORY_URL}
        - VITE_EASY_DATASET_URL=${VITE_EASY_DATASET_URL}
        - VITE_TENSORBOARD_URL=${VITE_TENSORBOARD_URL}
    networks:
      - docknet
  caddy:
    image: caddy:alpine
    ports:
      - "3000:3000"
    volumes:
      - ./Caddyfile:/etc/caddy/Caddyfile
      - caddy_data:/data
    depends_on:
      - backend
    networks:
      - docknet

  llama-factory:
    container_name: finetune-llama-factory
    build:
      context: ./llama-factory
      dockerfile: Dockerfile
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: all
              capabilities: [gpu]
    ports:
      - "7002:7860"
    volumes:
      - ./data:/app/shared_data
      - ./hf_cache:/root/.cache/huggingface
      - ./saves:/app/saves
      - ./models:/app/models
      - llama_factory_cache:/app/cache
    environment:
      - HF_TOKEN=${HF_TOKEN}
      - MPLCONFIGDIR=/app/cache
      - HF_HOME=/root/.cache/huggingface
    networks:
      - docknet

  easy-dataset:
    image: ghcr.io/conardli/easy-dataset
    container_name: easy-dataset
    ports:
      - "7001:1717"
    volumes:
      - ./data:/app/local-db
    restart: unless-stopped
    networks:
      - docknet

  llama-cpp:
    container_name: finetune-llama-cpp
    build:
      context: ./llama.cpp
      dockerfile: .devops/cuda.Dockerfile
      args:
        - CUDA_VERSION=12.8.1
      target: full
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: all
              capabilities: [gpu]
    volumes:
      - ./models:/models
      - ./data:/data
      - ./saves:/saves
      - ./hf_cache:/hf_cache
    entrypoint: []
    command: tail -f /dev/null
    networks:
      - docknet

  tensorboard:
    image: tensorflow/tensorflow
    ports:
      - "7003:6006"
    volumes:
      - ./saves:/app/saves
    command: tensorboard --logdir /app/saves --bind_all
    networks:
      - docknet

volumes:
  uploads:
  caddy_data:
  llama_factory_cache:

networks:
  docknet:
    external: true