LQDeployConfig/prod/models/minerU/compose.yaml

services:
  mineru-openai-server:
    image: mineru:latest
    container_name: mineru-openai-server
    restart: always
    profiles: ["openai-server"]
    ports:
      - 30000:30000
    environment:
      MINERU_MODEL_SOURCE: local
    entrypoint: mineru-openai-server
    command:
      --host 0.0.0.0
      --port 30000
      # --data-parallel-size 2  # If using multiple GPUs, increase throughput using vllm's multi-GPU parallel mode
      # --gpu-memory-utilization 0.5  # If running on a single GPU and encountering VRAM shortage, reduce the KV cache size by this parameter, if VRAM issues persist, try lowering it further to `0.4` or below.
    ulimits:
      memlock: -1
      stack: 67108864
    ipc: host
    healthcheck:
      test: ["CMD-SHELL", "curl -f http://localhost:30000/health || exit 1"]
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              device_ids: ["6"]  # Modify for multiple GPUs: ["0", "1"]
              capabilities: [gpu]

  mineru-api:
    image: mineru:latest
    container_name: mineru-api
    restart: always
    profiles: ["api"]
    ports:
      - 25428:8000
    environment:
      #MINERU_MODEL_SOURCE: local
       # 模型源：与 --source modelscope 保持一致
       - MINERU_MODEL_SOURCE=modelscope
       # 模型缓存路径（容器内）
       - MODELSCOPE_CACHE=/root/.cache/modelscope
       - MINERU_CACHE_DIR=/root/.cache/mineru
       # Transformers/HF 缓存，避免路径冲突
       - TRANSFORMERS_CACHE=/root/.cache/huggingface/transformers
       - HF_HOME=/root/.cache/huggingface
       # 日志与语言
       - LOG_DIR=/app/logs
       - LANG=zh_CN.UTF-8
       - PYTHONUNBUFFERED=1
    entrypoint: mineru-api
    command:
      --host 0.0.0.0
      --port 8000
      # parameters for vllm-engine
      # --data-parallel-size 2  # If using multiple GPUs, increase throughput using vllm's multi-GPU parallel mode
      # --gpu-memory-utilization 0.5  # If running on a single GPU and encountering VRAM shortage, reduce the KV cache size by this parameter, if VRAM issues persist, try lowering it further to `0.4` or below.
    volumes:
      # 1. 模型缓存持久化 (核心：避免重复下载)
      - /data/app_workspace/.cache/modelscope:/root/.cache/modelscope:rw
      # 2. MinerU 缓存持久化
      - /data/app_workspace/.cache/mineru:/root/.cache/mineru:rw
      - /data/app_workspace/.cache/huggingface:/root/.cache/huggingface:rw  # 新增：避免 transformers 缓存冲突
      # 3. 日志目录映射
      - /data/app_workspace/minerU/logs:/app/logs:rw
      # 4. 输入文件目录 (可选，如果 API 支持文件上传处理)
      - /data/app_workspace/minerU/input:/app/input:ro
      # 5. 输出结果目录 (可选)
      - /data/app_workspace/minerU/output:/app/output:rw
      # 6. 配置文件目录 (可选，如有自定义配置)
      - /data/app_workspace/minerU/config:/app/config:ro
    ulimits:
      memlock: -1
      stack: 67108864
    ipc: host
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              device_ids: ["6"]  # Modify for multiple GPUs: ["0", "1"]
              capabilities: [gpu]

  mineru-gradio:
    image: mineru:latest
    container_name: mineru-gradio
    restart: always
    profiles: ["gradio"]
    ports:
      - 7860:7860
    environment:
      MINERU_MODEL_SOURCE: local
    entrypoint: mineru-gradio
    command:
      --server-name 0.0.0.0
      --server-port 7860
      # --enable-api false  # If you want to disable the API, set this to false
      # --max-convert-pages 20  # If you want to limit the number of pages for conversion, set this to a specific number
      # parameters for vllm-engine
      # --data-parallel-size 2  # If using multiple GPUs, increase throughput using vllm's multi-GPU parallel mode
      # --gpu-memory-utilization 0.5  # If running on a single GPU and encountering VRAM shortage, reduce the KV cache size by this parameter, if VRAM issues persist, try lowering it further to `0.4` or below.
    ulimits:
      memlock: -1
      stack: 67108864
    ipc: host
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              device_ids: ["0"]  # Modify for multiple GPUs: ["0", "1"]
              capabilities: [gpu]