LQDeployConfig/dev/models/yaml_bak/docker-compose.yaml.nvidia

services:
  qwen3.6-27b:
    image: cr.metax-tech.com/public-ai-release/maca/vllm-metax:0.19.0-maca.ai3.5.3.502-torch2.8-py312-ubuntu22.04-amd64
    container_name: qwen3.6-27b-vllm
    shm_size: '10gb'
    ports:
      - "8004:30000"
    volumes:
      # # 宿主机路径:容器内路径
      - /opt/lq/models:/model:ro
      - ~/.cache/huggingface:/root/.cache/huggingface
      - /opt/lq/deploy_models/logs/logs:/var/log/vllm  # 日志目录映射
      - /opt/lq/deploy_models/bench_suite:/bench_suite #脚本目录映射
    environment:
      - CUDA_VISIBLE_DEVICES=1
      - PYTHONUNBUFFERED=1  # 确保实时输出
    command: >
      sh -c "mkdir -p /var/log/vllm &&
      python3 -m vllm.entrypoints.openai.api_server
      --model-path /model/Qwen3.6-27B
      --served-model-name Qwen3.6-27B
      --host 0.0.0.0
      --port 30000
      --tensor-parallel-size 1
      --max-num-batched-tokens 4096
      --max-model-len 8192
      --api-key sk-12345
      --log-level info 2>&1 | tee /var/log/vllm/qwen3.6-27b-server.log"
    ipc: host