---
notion-id: 23aca4d7-cb7b-808d-9918-e315643ea989
---
用于让 GPU 可被 Docker 容器访问：
```shell
distribution=$(. /etc/os-release;echo $ID$VERSION_ID)
```
```shell
curl -s -L https://nvidia.github.io/nvidia-docker/gpgkey | sudo apt-key add -
```
```shell
curl -s -L https://nvidia.github.io/nvidia-docker/$distribution/nvidia-docker.list | \
  sudo tee /etc/apt/sources.list.d/nvidia-docker.list

#cat /etc/os-release 查看版本不支持所以不运行下载

#手动将支持的版本（如 Ubuntu 22.04）写入配置，以绕过限制。
curl -s -L https://nvidia.github.io/nvidia-docker/ubuntu22.04/nvidia-docker.list | \
  sudo tee /etc/apt/sources.list.d/nvidia-docker.list

```
```shell
sudo apt update
sudo apt install -y nvidia-docker2
```
```shell
# 配置 Docker 默认 runtime
sudo tee /etc/docker/daemon.json <<EOF
{
    "default-runtime": "nvidia",
    "runtimes": {
        "nvidia": {
            "path": "nvidia-container-runtime",
            "runtimeArgs": []
        }
    }
}
EO
```
```shell
# 重启 Docker 服务
sudo systemctl restart docker
```

---
### (4) 验证 GPU 可用性
执行：
```shell
docker run --rm --gpus all nvidia/cuda:12.2.0-base-ubuntu22.04 nvidia-smi
```
若输出 GPU 型号和信息 ✅，即可用于 BLIP 模型部署。

---