rag-server/compose.yaml at master · aria3ppp/rag-server · 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
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
# yaml-language-server: $schema=https://raw.githubusercontent.com/compose-spec/compose-spec/master/schema/compose-spec.json

services:

  llm:
    image: ghcr.io/ggerganov/llama.cpp:server # this is the server latest tag
    command: ["-m", "/models/hugging-quants/Llama-3.2-1B-Instruct-Q4_K_M-GGUF/llama-3.2-1b-instruct-q4_k_m.gguf", "--pooling", "cls"]
    environment:
      LLAMA_ARG_PORT: 8081
    expose:
      - 8081
    ports:
      - "8081:8081"
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:8081/health"]
      start_period: 10s
      interval: 10s
      timeout: 5s
      retries: 5
    volumes:
      - ./models/:/models/:ro

  embedder:
    image: ghcr.io/ggerganov/llama.cpp:server # this is the server latest tag
    command: ["-m", "/models/CompendiumLabs/bge-small-en-v1.5-gguf/bge-small-en-v1.5-f32.gguf", "--embedding"]
    environment:
      LLAMA_ARG_PORT: 8082
    expose:
      - 8082
    ports:
      - "8082:8082"
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:8082/health"]
      start_period: 10s
      interval: 10s
      timeout: 5s
      retries: 5
    volumes:
      - ./models/:/models/:ro

  reranker:
    image: ghcr.io/ggerganov/llama.cpp:server # this is the server latest tag
    command: ["-m", "/models/BAAI/bge-reranker-base/bge-reranker-base-Q4_K_M.gguf", "--reranking"]
    environment:
      LLAMA_ARG_PORT: 8083
    expose:
      - 8083
    ports:
      - "8083:8083"
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:8083/health"]
      start_period: 10s
      interval: 10s
      timeout: 5s
      retries: 5
    volumes:
      - ./models/:/models/:ro

  qdrant:
    image: qdrant/qdrant:v1.12.4
    environment:
      QDRANT__SERVICE__HTTP_PORT: ${QDRANT_HTTP_PORT:-6333}
      QDRANT__SERVICE__GRPC_PORT: ${QDRANT_GRPC_PORT:-6334}
    expose:
      - ${QDRANT_HTTP_PORT:-6333}
      - ${QDRANT_GRPC_PORT:-6334}
    ports:
      - "${QDRANT_HTTP_PORT:-6333}:${QDRANT_HTTP_PORT:-6333}" # dashboard is available at http://localhost:${QDRANT_HTTP_PORT:-6333}/dashboard
      - "${QDRANT_GRPC_PORT:-6334}:${QDRANT_GRPC_PORT:-6334}"
    volumes:
      - qdrant:/qdrant/storage

  qdrant-healthcheck:
    image: busybox:1-musl
    entrypoint: ["sleep", "infinity"] # Keeps the service running forever
    depends_on:
      - qdrant
    healthcheck:
      test: ["CMD", "wget", "--no-verbose", "--quiet","--spider", "--tries=1", "http://qdrant:${QDRANT_HTTP_PORT:-6333}/healthz"]
      start_period: 10s
      interval: 10s
      timeout: 5s
      retries: 5

  rag:
    # image: ghcr.io/aria3ppp/rag
    # pull_policy: build
    build:
      dockerfile: ./docker/Dockerfile.rag
      context: .
    environment:
      RAG_SERVER_GRPC_PORT: ${RAG_SERVER_GRPC_PORT:-9001}
      RAG_SERVER_GATEWAY_PORT: ${RAG_SERVER_GATEWAY_PORT:-8000}
      RAG_SERVER_GATEWAY_ALLOWED_ORIGINS: ${RAG_SERVER_GATEWAY_ALLOWED_ORIGINS:-*}
      RAG_SERVER_GRACEFUL_SHUTDOWN_TIMEOUT: ${RAG_SERVER_GRACEFUL_SHUTDOWN_TIMEOUT:-30s}
      OPENAI_BASEURL: ${OPENAI_BASEURL:-http://llm:8081/v1}
      OPENAI_APIKEY: ${OPENAI_APIKEY:-apikey}
      OPENAI_MODEL: ${OPENAI_MODEL:-model}
      RERANKER_BASEURL: ${RERANKER_BASEURL:-http://reranker:8083/v1}
      VECTORSTORE_HOST: ${VECTORSTORE_HOST:-vectorstore}
      VECTORSTORE_SERVER_GRPC_PORT: ${VECTORSTORE_SERVER_GRPC_PORT:-9091}
    expose:
      - ${RAG_SERVER_GRPC_PORT:-9001}  # grpc
      - ${RAG_SERVER_GATEWAY_PORT:-8000} # http gateway
    ports:
      - "${RAG_SERVER_GRPC_PORT:-9001}:${RAG_SERVER_GRPC_PORT:-9001}" # grpc
      - "${RAG_SERVER_GATEWAY_PORT:-8000}:${RAG_SERVER_GATEWAY_PORT:-8000}" # http gateway
    depends_on:
      llm:
        condition: service_healthy
      reranker:
        condition: service_healthy
      vectorstore:
        condition: service_healthy
    healthcheck:
      test: ["CMD", "/app/rag", "-probe", "http", "-mute"]
      start_period: 10s
      interval: 10s
      timeout: 5s
      retries: 5

  vectorstore:
    # image: ghcr.io/aria3ppp/vectorstore
    # pull_policy: build
    build:
      dockerfile: ./docker/Dockerfile.vectorstore
      context: .
    environment:
      VECTORSTORE_SERVER_GRPC_PORT: ${VECTORSTORE_SERVER_GRPC_PORT:-9091}
      VECTORSTORE_SERVER_GATEWAY_PORT: ${VECTORSTORE_SERVER_GATEWAY_PORT:-8080}
      VECTORSTORE_SERVER_GATEWAY_ALLOWED_ORIGINS: ${VECTORSTORE_SERVER_GATEWAY_ALLOWED_ORIGINS:-*}
      VECTORSTORE_SERVER_GRACEFUL_SHUTDOWN_TIMEOUT: ${VECTORSTORE_SERVER_GRACEFUL_SHUTDOWN_TIMEOUT:-30s}
      EMBEDDER_BASEURL: ${EMBEDDER_BASEURL:-http://embedder:8082/v1}
      QDRANT_HOST: ${QDRANT_HOST:-qdrant}
      QDRANT_GRPC_PORT: ${QDRANT_GRPC_PORT:-6334}
      QDRANT_COLLECTION_NAME: ${QDRANT_COLLECTION_NAME:-collection}
      QDRANT_VECTOR_SIZE: ${QDRANT_VECTOR_SIZE:-384}
    expose:
      - ${VECTORSTORE_SERVER_GRPC_PORT:-9091}  # grpc
      - ${VECTORSTORE_SERVER_GATEWAY_PORT:-8080} # http gateway
    ports:
      - "${VECTORSTORE_SERVER_GRPC_PORT:-9091}:${VECTORSTORE_SERVER_GRPC_PORT:-9091}" # grpc
      - "${VECTORSTORE_SERVER_GATEWAY_PORT:-8080}:${VECTORSTORE_SERVER_GATEWAY_PORT:-8080}" # http gateway
    depends_on:
      embedder:
        condition: service_healthy
      qdrant-healthcheck:
        condition: service_healthy
    healthcheck:
      test: ["CMD", "/app/vectorstore", "-probe", "http", "-mute"]
      start_period: 10s
      interval: 10s
      timeout: 5s
      retries: 5

volumes:
  qdrant: {}