md/install-nvidia-smi.md

nvidia_cublas_cu12-12.8.4.1-py3-none-manylinux_2_27_x86_64.whl
## RTX 4090 笔电操作记录

```shell
# 因清华大学开源镜像站 HTTP/403 换了中科大的镜像站，配置信息存放在这里
cat /etc/apt/sources.list
# 安装显卡驱动和CUDA
wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2404/x86_64/cuda-keyring_1.1-1_all.deb
sudo dpkg -i cuda-keyring_1.1-1_all.deb
sudo apt-get update
sudo apt-get -y install cuda-toolkit-12-8
sudo apt-get install -y cuda-drivers
nvidia-smi
apt install nvidia-cuda-toolkit
nvcc -V
# 用 conda 创建了一个 python=3.12 的环境
conda create -n vllm-py12 python=3.12 -y
# 在该目录存放关于 vllm 和LLM相关的内容
cd /home/ss/vllm-py12
conda activate vllm-py12
export https_proxy=http://10.159.236.165:7890 http_proxy=http://10.159.236.165:7890 all_proxy=socks5://10.159.236.165:7891
pip install vllm --extra-index-url https://download.pytorch.org/whl/cu128
```