Linux/Ubuntu工具

使用vim添加环境变量

vim ~/.bashrc
# in vim normal mode, use G 可以到最底下
export PATH=/usr/local/cuda-11.8/bin:$PATH
export LD_LIBRARY_PATH=/usr/local/cuda-11.8/lib64:$LD_LIBRARY_PATH
source ~/.bashrc #保存更改

gdown下载Drive文件

1. 使用gdown（推荐）Link

pip install gdown
gdown [link] --fuzzy #for large file

3. 使用rclone cheetsheet
4. 使用wget/curl（实验未成功）

nvitop监控GPU使用量

1. 使用nvitop

pip install nvitop
nvitop -m full #full can be auto(default),full,compact

2. 使用watch监测

watch -n 1 nvidia-smi

df -h查看磁盘使用量

du -sh #查看当前路径下所有文件的大小
du -sh [folder/file] #可以是folder也可以是file，查看特定的大小
du -lh #tree form
df -h # shows the amount of disk space used and available on all currently mounted filesystems.

iftop监控网络流量

sudo apt-get install iftop
iftop -i eth0

一些正常配置

Cuda与CuDNN

# Download and install CUDA 11.8
wget https://developer.download.nvidia.com/compute/cuda/11.8.0/local_installers/cuda_11.8.0_520.61.05_linux.run
sudo sh cuda_11.8.0_520.61.05_linux.run --silent --toolkit

# Cleanup CUDA installer
rm -f cuda_11.8.0_520.61.05_linux.run

# Appending environment variable settings to ~/.bashrc
echo 'export PATH=/usr/local/cuda-11.8/bin:$PATH' >> ~/.bashrc
echo 'export LD_LIBRARY_PATH=/usr/local/cuda-11.8/lib64:$LD_LIBRARY_PATH' >> ~/.bashrc
source ~/.bashrc

# Download and install CuDNN
wget https://developer.download.nvidia.com/compute/cudnn/9.0.0/local_installers/cudnn-local-repo-ubuntu2204-9.0.0_1.0-1_amd64.deb
sudo dpkg -i cudnn-local-repo-ubuntu2204-9.0.0_1.0-1_amd64.deb
sudo cp /var/cudnn-local-repo-ubuntu2204-9.0.0/cudnn-*-keyring.gpg /usr/share/keyrings/
sudo apt-get update
sudo apt-get -y install cudnn
# To install for CUDA 11, perform the above configuration but install the CUDA 11 specific package:
sudo apt-get -y install cudnn-cuda-11
rm -f cudnn-local-repo-ubuntu2204-9.0.0_1.0-1_amd64.deb

# 可选项：使用deb方式安装CUDA会自动安装gnome组件，对于无图形界面的服务器来说这些组件是多余的，可以通过以下命令删除。
sudo apt remove --purge gnome-*
sudo apt autoremove --purge

检查Cuda和Pytorch版本。test.py

一些杂碎配置

glibc更新 Official Link（未尝试）

Some command

strings /usr/lib/i386-linux-gnu/libstdc++.so.6 | grep GLIBCXX # show all glibc version
ldd --version 

sudo apt install bison
sudo apt install gawk

Nvidia Driver安装 for ubuntu Link (未尝试)

Vast.ai传数据

使用Google Drive导入Instance例子 path to destination的路径使用pwd命令得到

GCP(Google Cloud)

---